Table of Contents
-subj
of openssl req/dev/fd0
List of Figures
List of Tables
List of Examples
ldap.xml
ldap.xml
/etc/mail/sendmail.mc
batch_new_user.txt
batch_new_process.txt
batch_process_data.txt
Public Key Infrastructures (PKIs) are one of the most widely accepted musts of the future. The problem is that most applications can be secured with certificates and keys but it is really difficult and expensive to setup PKIs, the reason being that flexible trustcenter software (especially for Unix) is expensive. This was the starting point of OpenCA. Our goal is production of an open source trustcenter system to support the community with a good, inexpensive and future-proof solution for their base infrastructure.
OpenCA started in 1999. The first idea consisted of three major parts - a Perl web interface, an OpenSSL backend for the cryptographic operation and a database. This simple concept is still our motto today. Nearly all operations can be performed via some web interface. We have six pre-configured interfaces and many more can be created from them, depending on the need. The cryptographic backend is still OpenSSL, which is in no way a disadvantage. We want to build the organizational infrastructure for a PKI. This is our major goal and the guys from OpenSSL have much more experience with cryptography than we have. Our databases store all the needed information about the users' cryptographic objects like Certificate Signing Requests (CSRs), Certificates, Certificate Revocation Requests (CRRs) and Certificate Revocation Lists (CRLs).
If you imagine that the development of OpenCA will be finished in some weeks or months then you are probably on the wrong way. There are many items which still need to be implemented. Today we support the following elements (this is an incomplete list just to give you an impression of how complex the subject matter is):
Public interface
LDAP interface
RA interface
CA interface
SCEP
OCSP
IP-filters for interfaces
Passphrase based login
Certificate based login (including smartcards)
Role Based Access Control
Flexible Certificate Subjects
Flexible Certificate Extensions
PIN based revocation
Digital signature based revocation
CRL issuing
Warnings for soon to expire certificates
support for nearly every (graphical) browser
OpenCA is designed for a distributed infrastructure. It can, not only handle an offline CA and an online RA, but using it you can build a whole hierarchy with three or more levels. OpenCA is not just a small solution for small and medium research facilities. The goal is to support maximum flexibility for big organizations like universities, grids and global companies.
The OpenCA guides consist of four parts: The first part is a design guide which should help you to setup a good infrastructure. The second part describes all the activities which must be performed offline by some administrator. The third part is the user guide which describes all the available features. The last part is the technology guide which documents the ideas behind OpenCA. This part is meant only for developers and hardcore administrators to understand what's going on.
Finally we wish to thank everybody who helped us program, test and document OpenCA. This also includes of all the universities and companies which finance the work of our developers.
Table of Contents
The meaning of this guide has changed. It must be reorganized. It should now help people to implement a PKI which use the full power and flexibility of OpenCA. Mainly it should give the people an overview about the hierarchies which can be implemented with OpenCA.
Before you start installing or editing OpenCA you should understand the most important principles of OpenCA's design. The first section describes the general design and the second section describes the software design.
This guide will be incomplete at every time because it is the part of documentation which must be changed at every time the software changes. So please be lenient toward us :)
Table of Contents
We start here from scratch to give everybody a chance to understand how OpenCA works. So if you think about these boring guys who write this, please take in mind that OpenCA novices must also have a chance to understand the software.
The basic idea of every X.509 PKI (Public Key Infrastructure) is a strong hierarchical organization. This results in a tree of databases if we try to create a distributed PKI architecture.
The data exchange between such isolated databases can be handled automatically if you use a distributed database system but in the sense of OpenCA such a distributed database system is only one database in our tree. If you really have an isolated database (e.g. for an Offline CA) then you must have the technology for the data exchange and the management of the complete node in the hierarchy. This management functionality is bundled in an interface called node or node management. Hence the design of OpenCA looks like follows
Normally every server in the infrastructure of the trustcenter has it's own database for security reasons. This hierarchy is the backbone of the trustcenter.
After you know the basic infrastructure of OpenCA you possibly want to know what we think about such things like CA, RA, LDAP and a public interface which is sometimes called web-gateway? OpenCA supports all these software components via special web interfaces.
If you want to design a powerful trustcenter then you must have a concept about how you want to organize your work flow. You can see an example in the following figure.
OpenCA actually supports the following interfaces:
Node (for node management)
CA
RA
LDAP
Pub
SCEP
This interface manages the database and handles all the export and import functionalities.
The database can be initialized what means that OpenCA can create all the tables but OpenCA cannot create the database itself because this differs for every vendor. So we need a database with the appropriate access rights and a new database. The interface includes some functions for the backup and recovery of such a node but please bear in mind that you MUST have a separate backup of the CA's private key and certificate. There is no default mechanism in OpenCA to backup the private key. We don't implement it because first we found no general secure way to backup a private key and second the most CA's use HSMs and therefore you need a completely different and usually proprietary backup strategy.
The export and import will be handled by this interface too. You can configure different rules for the synchronization with nodes on a higher and a lower level of the hierarchy. This includes the configuration of the objects and status which can be exchanged. The configured filters avoid status injections from lower levels of the hierarchy.
The CA interface has all the functions which you need to create certificates and Certificate Revocation Lists (CRLs). The CA also includes all the functions which you can use to change the configuration via a web interface. It is not possible to change the configuration via another web interface.
The CA is the home of the batch processors too. OpenCA includes some powerful batch processors for creating certificates. These batch processors can be used for automatic certificate creation from various Enterprise Resource Planning (ERP) systems (e.g. SAP, HIS, NIS or /etc/passwd).
OpenCA's RA is able to handle all kinds of requests. This include things like editing requests, approving requests, creating private keys with smart cards, delete wrong requests and email users.
The LDAP interface was implemented to separate the LDAP management completely from the rest of the software. This is necessary because there are many functions which are really specific for LDAP admins, with only a few users needing these features.
The Public interface includes all the small things which the users need. This is only a small list and perhaps it is incomplete
generates CSRs (certificate signing request) for Microsoft Internet Explorer
generates CSRs for Mozilla 1.1+ and Netscape Communicator and Navigator
generates client independent requests and private keys (e.g. for KDE's konqueror or server administrators who don't know how to create a private key and request)
receives PEM-formatted PKCS\#10 requests from servers
enrolls certificates
enrolls CRLs
supports two different methods revocation
search certificates
tests user certificates in browsers (Microsoft Internet Explorer and Netscape Communicator and Navigator 4.7x)
Sometimes there is a confusion about the status of OpenCA objects. In the following figure you can check here the complete life cycle of all OpenCA objects. If this figure is not complete, or if you find a mistake or you don't understand something then please contact openca-users@lists.sf.net
Table of Contents
This chapter should give you a lot of hints which you should bear in mind if you design your first PKI. Please don't ignore this section if you are an experienced PKI administrator. We try to list the big traps here. So if you know another major problem then please add it.
This section lists hardware issues which were a problem for some PKIs during their production use. This list does not discuss performance issues.
One of the biggest problems of PKI systems is the time. There are two different kinds of computers - online and offline. The usual administrators logic is that a network connected computer can use a timeserver. The question is can you trust this timeserver? A timeserver creates two problems. First is the timestamp really from the timeserver and second is the time source of the timeserver trustworthy? The connection to a timeserver can be secured via tunnel technologies like IPsec but the real problem is the time source. The most timeservers finally use a radio clock which receives a unsigned time signal from a radio station. This signal can be easily faked because it is very weak. So network time sources are really insecure.
After discussing the disadvantages of using online computers we can discuss offline technologies. Radio clocks are problematic and hence we need not discuss them twice. Also many buildings made from good ferroconcrete do not have problems with radio signals because they act as really nice Faraday cages! So what are the alternatives? First we need a trustworthy time source. Simply take a digital watch and compare it's time with several other clocks on the Internet, the video text of your TV, a radio clock, the sun ;-), GPS and any other source you can find. Second you transfer the time from your watch to the computer. Last but not least you have to check the drift and the clock itself on the computer. The drift is a small and easy to handle problem. The clock itself is a much bigger problem. If your computer is always connected to a power outlet then the clock should only drift. Please remember this if you put your CA on a notebook and the notebook into a safe. Several new notebooks have really bad CMOS batteries which result in a wrong time at every reboot. So as you can see time is not trivial.
The most common hardware crashes involve cooler and disk failures. You should backup all your important data - especially ALL issued certificates. Never lose a certificate or you must revoke the complete CA. Backups are a nice thing but it costs some time to recover from a backup. This results in two facts. First you must have a detailed (time-) plan how to recover from a backup. Second you should be able to tolerate disk crashes. RAIDs are sometimes expensive but they helps a lot (ask your SAN admins :) ).
Usually you can make a visual observation if your laptop crashes. A crashed offline computer can be detected by visual monitoring too. However a crashed online component of a PKI is problem because important information is not available,such as newly issued certificates and CRLs. Such a situation also brings down your services like SCEP. If a public interface of the security infrastructure is down then you are bound to have a trust problem with your users in the future. It is also noteworthy that a simple ping is not enough to detect hardware failure. You cannot detect a crashed web or OpenCA perl server with a ping. Software bugs can also cause 100 percent load. I know this problem from our web mail programs really well :)
If your offline CA consists of one offline notebook and you want to ensure dual control and no single point of failure then you can the following method as an example: use one IT module safe and two data safes. An IT module safe is a safe with its own climatization and UPC which has the same physical protection level like a safe. This safe is used to allow nonstop operation of the notebook which reduces time and availability problems. All three safes should have two locks. This ensures dual access control by key sharing. The setup is really simple and really efficient too.
The organization of the safes is really easy. You split the CA passphrase into two parts - front and back. The organization of the data and computers is like follows:
The IT module safe includes the notebook (with the CA private key) and the front part of the passphrase.
The first data safe includes the backups (including the private key backups) and the back part of the passphrase.
The second data safe includes the front and back parts of the passphrase.
This organization ensures that one broken safe doesn't corrupt the infrastructure. It is important to immediately start a roll out of a new infrastructure but there is no reason for panic. This arrangement also ensures that a loss of one safe doesn't stop your operation. You need two safes to survive and the loss of one safe is acceptable at minimum for a short period of time.
Please remember that this is a really simple idea for medium risk CAs. High risk CAs should use more complex schemes to tolerate more than one broken safe. They should be able to tolerate at least two broken safes to have a longer schedule for the roll-out of a new CA.
This is more an area for a facility manager. The rooms with the PKI safes inside should have solid walls and a door with two locks. The room including the climatization system should be fire proof for 90 minutes (F90). The room and the entry should be camera observed. The cameras in the room itself should show the persons but not the keyboards and monitors. Papers should not be readable. The recorder for the camera should record one week at minimum. The room should also have an alarm system and must be safe against electro magnetic pulse (EMP) and water flooding. This is only a short summary of ideas. Please ask some assurance specialists or architects for more details.
A PKI is only fully functional if all the services provided by it are fully operational. This include not only things like OCSP and SCEP but the public gateway as well. Many people think it is enough if the OCSP and one CDP keeps working but this is not sufficient. The first reason being that most applications don't understand OCSP. The second problem is that the last running CDP might only support LDAP, regardless of the fact that there are applications which only support HTTP. Even more problematic is a single running HTTPS CDP. The core mistake in the assumption is the meaning of fully operational. A PKI is not fully operational if only the CDPs still work. Nobody can download a new certificate or the certificate of an existing user in such a state. The PKI would still be secure but it would not be operational!
In a time of server consolidation and omnipresent networks it is important to understand that all public PKI services must be available after a single failure. This includes network and power outages. A second fiber only helps if it is not in same pipe like the other one. Digging crews don't differentiate between the fibers if they accidentally cut a pipe! I know this situation really well :( If you have big distributed unit then it is recommended that at the minimum two of these units run the public interface. In this case you should also have independent interconnects.
This section consists of two big groups of problems - CDPs and application specific problems. Applications are also discussed in other sections of the OpenCA guide in case we have developed special configurations to solve the problem.
CRL Distribution Points are a critical area. There were several PKIs in the earlier years which had no CDPs or only one CDP. Secure applications must verify the state of a used certificate. Such applications check the CDP field in the extensions of a certificate to find a usable verification source. Today a CRL Distribution Point should not be a source for a CRL. It can be an OCSP responder. A CDP is today more of a Certificate Status Point.
The first question is which protocols must be supported. The common protocols LDAP and HTTP should be supported always. HTTP is supported by nearly all devices but some devices prefer LDAP over HTTP - especially some network solutions. Additionally there are OCSP and HTTPS. OCSP is a protocol to verify the state of a single certificate. This protocol has a much better performance for slow network interconnects if you have large installations with many (revoked) certificates. HTTPS supports you with a trusted source, but is a trusted source necessary for CRLs? Usually not but sometimes yes. CRLs are signed by the CA and they have period of validity. There is only one working attack for CRLs. If a certificate was revoked, but the old CRL is still valid and if the CDP uses HTTP, then an attacker could present the user the old CRL. So this attack is more a question of security policy than a technical question. The validity period is a policy question too because most applications don't accept CRLs if the timestamp for the next update is over. The cool thing is that a CRL is never invalid! There can only be a newer CRL but as mentioned earlier this is more a political discussion.
The second question is how many physical CDPs do you need. If you design a scalable infrastructure then you must be able to tolerate service interruptions of several components like cables, routers, switches and servers. It is a really good idea to always have a local CDP if you only have a single interconnect to your central network. However not every branch office needs its own CDP, since if their computers (including their file and mail systems) are offline then it is not necessary to have a CDP online.
The third question is tricky. Which order should the CDPs have? The question is interesting in the case of network outages. If you support LDAP and HTTP, you have three servers which all serve both protocols. The CDP are ordered like the servers and servers one and two are not available because of a more or less big digger then your applications waits for four timeouts until it can access the first CDP successfully. The application has to wait for timeouts from LDAP on server one, HTTP on server one, LDAP on server two and HTTP on server three. So it is a really good idea to mix the order of protocols and servers. This is most important for application which understand many protocols. Stupid alphabetical ordering using the DNS names is not recommended. Do you think a user would wait two minutes per recipient of an email, for verification if he wants to send a commit for a meeting?
The last question often, does not have a complete solution today. Can some supported and necessary protocols crash other applications? If you use HTTPS then you should invest some time on this problem because it is a typical chicken and egg problem. If you have Microsoft client and the client tries to verify the certificate status via a HTTPS CDP then it receives a certificate from the HTTPS server. This certificate must be verified. If the certificate is from the same CA then the client would contact the HTTPS CDP again to verify the certificate and the verification problem would start again. This is a nice method to implement a busy wait. The only solution is to get a server certificate from another authority which use no CDPs or at minimum no HTTPS CDPs. The insight is that you should be really careful with the usage of crypto protected CDPs.
If you start thinking about email security then the starting point is S/MIME or PGP. However at times ethereal admins just use IMAPS and POPS. So far so good. Later they think about server based security and try to implement SMTP over TLS. Certificates issued by OpenCA can be used for SMTP over TLS because the certificates can act both as a client and a server certificate. If you use some other software than OpenCA, or you create a new role for such servers then you should read the documentation of your SMTP server really carefully.
A SMTP server can use two methods for SMTP over TLS. It can use a single certificate which includes the extensions for TLS client and servers, or it can use two certificates - one client and one server certificate. Please read the specs for the certificates carefully and then read specifications for TLS clients and TLS servers. If the SMTP server reports an error after start tls then there is usually a missing extension or a problem with the certificate chain. See Section 3.2.2, “SMTP server”.
Old Netscape clients do not conform with RFC standards. They use the common name as a regexp to match the server name and ignore the subject alternative name completely. A workaround is described in the configuration area of OpenSSL (see Section 3.2.1, “HTTPS server”).
OpenLDAP is an open source software but it's TLS implementation is not the best one. It doesn't check the subject alternative name first, instead it checks the common name of the subject to match the DNS name or IP address. This fails if the certificate subject includes a regexp for old Netscape clients. Two non-standard compliant software packages collide here. Netscape needs regular expressions, OpenLDAP does not support regular expressions and both software packages do not check the subject alternative name first. So it's your job to make a trade-off, or to use Mozilla!
dual access control (physical and technical but first organizational aspects)
access control but who controls the access control regulations (no self control)
how to integrate privacy officers public certificates - not always public certificates - which fields must be published? PID vs. new ID what is the identity of a person in conventional areas?
There are numerous situations where it is a good idea to operate more than one PKI for end users. Perhaps you need a server CA and a separate user CA. Sometimes an old CA is still active in issuing CRLs because there are still valid certificates but the new CA issues the newer certificates. Other people use different CAs to establish an easy access control by certificate chains (so called trust paths). As you can see there are really many situations where you have to operate more than one PKI.
Most PKI programmers like me have no problem in distinguishing between different PKIs, because we always ask ourselves as to who issued the certificate. Normal users instead look at the certificate, call the hotline and ask why their certificate for Jon Doe with serial 12345 does not work. At the other end the guy from the hotline looks into his computer and answers that the certificate is correct and valid. So what's going on?
A certificate has two significant elements to identify a certificate, which are different from the common name in the subject of the certificate, and which are easy to handle. The keyID and issuer from the authority key identifier are however not easy to interpret for an end user. First there is a serial and second there is an issuer. If a customer calls a hotline then the easiest way to handle a problem is by using organization wide unique serial numbers. If you start a second CA or you have to replace an old CA, never reuse serial numbers if possible. You will have to search for hours if somebody calls you and reports a broken certificate chain for a certificate 12345 when you have two of those certificates. If you ever issued certificates with identical serials then always ask the issuer if you receive an error report. Never ever create a replacement for an old CA with the same name. It only causes trouble.
To sum it up in a simple manner: If you avoid duplicate identifiers then you automatically avoid many problems.
Table of Contents
This guide should describe all installation and administration issues of OpenCA. Some answers are perhaps in the FAQ but every detail which you can find in our docs about the istallation you can find here.
OpenCA is not a complete monolithic system. It uses several software products from other developers of the Open Source community. The following things are used:
Apache
mod_ssl
OpenSSL
OpenLDAP
Perl
We use a lot of different Perl modules. Beginning with OpenCA 0.9.2 we no longer install all foreign modules. This is the normal behaviour of every Open Source project. The following should give you an overview about the required modules. Please note that you must install at minimum the listed version because some earlier versions like for example Net::Server include serious bugs.
Table 3.1. External Perl modules
Module | Version | Comment |
---|---|---|
Authen::SASL | 2.04 | required by Net::LDAP for SASL authentication - if you do not use SASL then you do not need it |
CGI::Session | 3.95 | required for our own session handling |
Convert::ASN1 | 0.18 | ??? |
Digest::HMAC | 1.01 | required by Authen::SASL |
Digest::MD5 | 2.24 | this is usually part of Perl itself |
Digest::SHA1 | 2.02 | required by OpenCA itself |
Encode::Unicode | ??? | required by OpenCA for the internationalization stuff |
IO::Socket::SSL | 0.92 | ??? |
IO::stringy | 2.108 | ??? |
MIME::Base64 | 2.20 | required for Base64 encoding and decoding |
MIME::Lite | 3.01 | required for OpenCA mail handling |
MIME-tools | 5.411 | required for OpenCA mail handling |
MailTools | 1.58 | required for OpenCA mail handling |
Net-Server | 0.86 | required for OpenCA daemon - the version is important |
Parse::RecDescent | 1.94 | required by X500::DN |
URI | 1.23 | ??? |
X500::DN | 0.28 | we use a modified version here |
XML::Twig | 3.09 | used for XML parsing
Warning
Please read the file |
libintl-perl | 1.10 | this is our interface for the i18n stuff |
perl-ldap | 0.28 | Perl's LDAP interface |
OpenCA was tested on several softwarearchitectures but not on so many hardware architectures. Therefore we publish a list of used hardware. Please remember that OpenCA can be used on any system which support Apache, mod_ssl, OpenSSL and Perl. So if you have Unix box then it is usually possible to run an OpenCA on it.
i386 with Linux, FreeBSD, OpenBSD and NetBSD
UltraSparc with Solaris 8 and Linux
PowerPC with AIX
OpenCA uses the usual Open Source method to configure the source. We only use configure to compile and install the software but we don't use configure for the configuration of the installed system. configure make some defaults settings but the real configuration is described in the post-install section.
We will describe the ideas and options in the next section grouped by
such things like path settings, mail, web-server related stuff.
If you don't understand an explanation then please contact
<openca-user@lists.sf.org>
. The install options are
now lesser because we changed the installation process from
0.9.1 to 0.9.2 to get usable packages and better internationalization.
We don't document the general options of configure because it is not our job to document autoconf. We will only describe OpenCA specific options.
You should define the used system before you start configureing OpenCA itself. OpenCA must know several parameters about your system to work properly.
Table 3.2. Supported parameters for host configuration
Parameter | Description |
---|---|
--with-openssl-prefix =DIR
|
Usually OpenSSL is present on the most Unix systems because it is the best available Open Source cryptotoolkit. The problem is that several old distributions only include support for OpenSSL 0.9.6 but OpenCA needs version 0.9.7.
If you install an OpenSSL from source then it installs
in |
--with-openca-user =ARG
| OpenCA installs several files which should not be owned by the webserver user. Usually the owner can be root or special OpenCA user. It is recommended to use another user than root. |
--with-openca-group =ARG
|
OpenCA installs several files which should not be owned
by the webserver group. Usually the group can be root or
special OpenCA group. It is recommended to use another
user than root. If you install several CA you can setup
a group openca or pki for
example .
|
We have three different groups of paths - common stuff, prefixes for the different components of OpenCA and the paths for files of the webserver.
One path cannot be classified -
--with-module-prefix
=DIR
.
This path can be used to put all Perlmodules which OpenCA
installs in one directory to be able to remove OpenCA from your
system without any residues. It is also a good idea to use this
option if you need different OpenCA installations with
different versions of OpenCA on your system. Later versions of
OpenCA can have different modules with different interfaces
which are not backwards compatible.
OpenCA includes a directorystructure to store all relevant
data in one central place. This place can be specified with
--with-openca-prefix
. This installation
option is recommended for normal installations from the source
code. Secure or not the most users want to install packages
(e.g. RPM or DEB). Packages have the big advantage that you
remove or add a software without any risks. In this case we
have to support the package maintainers with configuration
options to build packages which conform with the guidelines
for the distros. Therefore you can use
--with-etc-prefix
,
--with-lib-prefix
and
--with-var-prefix
too.
Today there are six different components - ca, ra, ldap, pub, node and scep. Every component must have a different name to have distinguished configuration files and distinguished paths. All the names will be calculated automatically. You have only to edit these prefixes if you need a special configuration like a second RA on the same machine.
The webserver configuration is the most complex and most simple part of the configuration too. If you have single http-server for OpenCA then you only need four options to configure OpenCA for this server. If you have a full featured corporate portal then you can integrate this software seemlessly in the the server. Therefore you can configure a lot of details. So we hope you find a good tradeoff ;-)
Every webserver needs some basic informations. These
informations are the hostname
(--with-web-host
), the user
(--with-httpd-user
) and the group of the
server (--with-httpd-group
). These are the
rudimentary informations which OpenCA needs before you can start
configuring the paths. The defaults are
an empty hostname, nobody
and
nogroup
.
The most trivial installation case is the default apache
installation. In this case you have only to set
--with-httpd-fs-prefix
to the directory where
your apache is. All other directories will be set automtically.
The standard webserver doesn't use Apache's default
installation. Therefore it is possible to configure every detail
of the installation. The first splitting is into CGI
(--with-cgi-fs-prefix
) and HTDOCS
(--with-htdocs-fs-prefix
). The most test
systems don't need the other options. They have only to know
where the appropriate directories are.
Our software was designed for really big companies and
organizations too. They have usually portals for their employees
and customers. If you have to integrate an OpenCA interface into
such a portal then there are good news for you - you don't have
to edit paths and links by hand. You can configure the placement
of CGI and HTDOCS area of every interface seperately. The
options are
--with-(ca|ra|ldap|pub|node|scep)-(cgi|htdocs)-fs-prefix
).
We think that more flexibility is not necessary. So if you think
OpenCA is to unflexible then write a mail to us with your ideas.
The mailoptions are deprecated too. Please read the post-install section to understand how to configure mail. Please don't use the configure option because they can be removed in the next releases.
You can enable three extra features for compilation and
installation. SCEP and OCSP can be enabled because they are extra
softwarepackages which can work independently from OpenCA but they
are included in the distribution. The option
--enable-package-build
is used to support
package maintainers. If it is activated all common parts of OpenCA
are not installed automatically. This allows packagers to build
seperate conflict free packages for every interfaces because all
Perl modules and the common stuff can be put into seperate packages.
First run make, second run make test and then run the different install commands. make ca and make ext are the same like make You have the following install options for make.
install-ca
install-common
is only interesting for package
maintainers because they can install the common stuff seperately
from the rest.
install-ext
is the same like install online.
This install target is deprecated.
install-ldap
install-node
install-offline
installs
ca and node
install-online
installs
ra, ldap,
pub, scep and
node
install-pub
install-ra
install-scep
install-docs
OpenCA incudes a startup script for it's daemons. The script is named
OPENCADIR/etc/openca_start
. The script starts the
XML cache and the main server loop of OpenCA. Please remember to run
this script after every boot operation. It is recommended to integrate
a script openca
to the appropriate runlevel.
After the installation all necessary files are in the correct
directories but there are hunderts of files called
*.template
. These files contain placeholders which
can be configured in OPENCADIR/etc/config.xml
.
Before you start using OpenCA check this file and run
OPENCADIR/etc/configure_etc.sh.
OPENCADIR/etc/configure_etc.sh loads
OPENCADIR/etc/config.xml
and creates the correct
files. If you use packages from distribution then OPENCADIR is usually
empty because they create a directory /etc/openca
.
config.xml
contains seven big sections which will
be described first. Second we describe how to setup an installation with
only one common area but two management interfaces. This is useful if
you want to test the dataexchange.
Here you have to define some options which are relevant for
several interfaces. The ca locality, organization and country
affects the distinguished name and the preconfiguration of the LDAP
stuff. Nevertheless you should read the LDAP section too. The values
which you enter are directly used for l
,
o
and c
.
The mailpart is used for the node and RA interface. The
sendmail
field defines the command which will be used to send mails. You must
have a mailprogram with a sendmail interface (e.g. postfix). You can
enter every program which works like sendmail -n.
There are several people which like postfix more than sendmail and
we don't like do decide which mailprogram is the best one. The
option send_mail_automatic
configures the node
interface. If the value is YES
then OpenCA
sends all incoming mails during an import automatically. This can be
nice but it is dangerous too if you make a mistake. The
service_mail_account
is used as
From
for all sended mails. Usually this should
be something like
<Registration Authority <pki@your_org.edu>>
.
You can replace >
by
>
to but this is not required by the
XML specification.
The last option policy_link
defines a link to
your policy. It is highly recommended to don't ignore this value.
If you have such a reference then you can modify the page
request_success.html (add a hint) and the users can read all about
the PKI at every time they want. If you have no such link then you
receive dozens questions which are really simple but cost a lot of
time and you have no base for your operations. Ok, I think I have
not to explain the advantages of a policy here ...
Sometimes you need to run a CA on really unusual ports or you have
to use https. It is also a little bit difficult for us to guess
your correct hostname. Therefore you can specify these parameters
in config.xml
. The httpd_port
should be the default port of the protocol and in this case it can
be empty. If you need to run for example a http server on port 8080
then you have to use the option. Please remember to set the colons
if you specify a port.
CRL distribution points (CDPs) can be specified extra. This is necessary because there are softwares which has problems with https in general or other softwares which try to download a CRL and before they start the download they want to check the CRL of the webserver but the CRL is not present (or why should somebody tries to download it :) ) and so an endless loop starts - Microsoft CAPI is such a software.
If you setup a real CA then it is highly recommended to
edit all files in OPENCADIR/etc/openssl/extfiles
and OPENCADIR/etc/openssl.cnf
too. Every
certificate should contain at minimum two CDPs. It is best practice
to have two http CDPs and two ldap CDPs. Such a solution allows fast
migration and a good reliability.
Before you start working with OpenCA's LDAP code please be sure
that your LDAP server knows the objectclasses
pkiUser
, pkiCA
and
uniquelyIdentifiedUser
. The last objectclass
was introduced by Entrust Technolgies to have a clean way to include
serialnumbers into the subject of the certificate. Yes, it is
proprietary but there is no other way to do it.
The following list is identical with the listof the list in the tech guide where you can find more informations about OpenCA's LDAP code and how to configure the details in the configuration files.
If you set this option to "yes" then the LDAP code will be activated.
If you want that the LDAP server will be updated during
the import from a higher level of the hierarchy then
you must set this option to YES
.
This is the hostname of your LDAP server.
This is the port where your LDAP server listens.
The bind DN of the user which OpenCA uses to add data to the server.
The passphrase for OpenCA's ldap account.
First you have to decide which database module you want to use.
OpenCA supports two modules - one for DBM files and one for SQL
databases. DB
activates support for DBM files
and DBI
activates the SQL support.
If you want to use SQL databases then you have to setup some additional parameters:
This is the type of the DBD driver. We support Pg, MySQL, Oracle and DB2. If you need support for another database then please contact us.
name of the database which OpenCA should use
host of the database but sometimes the drivers don't need the host.
same as for host
the database user for OpenCA
has not to be explained :)
OpenCA has a mechanism to isolate the different interfaces from
eachother to avoid conflicts especially double serialnumbers.
The module_shift
is the number of bits reserved
for the IDs. You can use IDs from 0 to
(2^module_shift
- 1). 0 is the ID of the CA.
All the other _module_id
s must be in the scope
of the module shift. Please be careful you cannot change the
module_shift
after the first definition.
The _url_prefix
options define the exact
positions in the webserver. This depends highly on the
positions of the files in the filesystem but you can configure
aliases in the httpd.conf
. So OpenCA
is fully flexible.
SCEP is really simple to configure but please don't forget the access control configuration. It is strongly recommended to restrict the source addresses which can access the SCEP server.
This is the PEM encoded private key of the SCEP interface. It has the same format like for mod_ssl.
This is the PEM encoded certificate of the SCEP interface. It has the same format like for mod_ssl.
This is the passphrase for the private key of the SCEP server. If you use a not encrypted private key (what is not recommended - then please set an empty string here. interface. It has the same format like for mod_ssl.
The configuration of the dataexchange is really complex in OpenCA. You can find a description in the section about the configuration of the dataexchange (see Section 9, “Dataexchange”). If it is your first OpenCA installation then please use one of the templates. If you setup a production level PKI then you must understand this configuration before you use it. This is one of the most important configuration options to guarantee the security of the PKI.
Before the explanations start please notice that it is important to first install the online components and then the offline components if you follow the instructions because the configuration of the offline components take care about the already configured online components.
Additonally please remember to set the configure option
--with-node-prefix
to different names
during the configuration of the online and offline
installation. This is important because otherwise you have
only one management interface.
The first installation uses only the normal steps -
./configure --with-node-prefix=online_node --with-your-options,
make, make test,
make install-online,
edit OPENCADIR/etc/config.xml
and
OPENCADIR/etc/configure_etc.sh.
Please use your options to configure the software and use the
hierarchy level ra
.
The first step is the protection of the already installed
configuration files. Please set no permissions to the later
needed configuration files in
OPENCADIR/etc/servers
.
chmod 000 etc/servers/*.conf*
The first four steps are the same as for the online components
except of the configuration options where you should change at
minimum the hierarchy level to CA
.
So first you do
./configure --with-node-prefix=offline_node --with-your-options,
make, make test,
make install-offline and
edit OPENCADIR/etc/config.xml
.
The next step is really important you have to edit the file
etc/configure_etc.sh
. The directory with
the serverconfigurations is protected because of the first step
but all the other directories should only contain configuration
files of the ca. Usually there should be the following
directories:
/Test/OpenCA/etc/ /Test/OpenCA/lib/servers/offline_node /Test/OpenCA/lib/servers/ca /Test/htdocs/ca /Test/htdocs/offline_node
After you fixed the script please run it. Now the offline components are installed and configured.
menu.xml
must be fixed manually because it
includes only a basic configuration. You have to copy a
complete menu
section. The section must be
renamed from offline_node
to
online_node
. The cgi_prefix
must be fixed too. Please verifies the menus with the
name
s ra
,
ldap
and pub
to
use the correct links to the node interface. If all values
are correct then you have now a working testinstallation with
two management interfaces.
Table of Contents
Before we begin with the access control itself some
introductorily notes about the module ID.
Every module in OpenCA has a module-ID. This ID you can find in the
configurationfile of a module. The ID is used to create unique serial
numbers for the requests. The moduleshift defines how many bits at the
beginning of a serial number (the least significant bits) are reserved for
the module's ID. The advantage is that you can issue a request at any
module of OpenCA without synchronizing the databases at every time you
issue a request because the module's ID is part of every request serial.
The parameter in the configurationfiles are ModuleID
and ModuleShift
. You can configure both parameters via
./configure. The options are
--with-module-shift
,
--with-ra-module-id
and
--with-pub-module-id
. The ID of the
CA is at every time zero.
OpenCA includes since version 0.9.2 a very powerful access control. The
configuration of this access control is completely XML based.
The configuration files are placed in
OPENCADIR/etc/access_control
.The
document root element is openca and the complete configuration is placed
in the area of the access_control element.
<openca> <access_control> The complete configuration should be here. </access_control> </openca>
The complete system consists of four parts:
channel verification
login
session management
ACLs
Every step is a completely isolated pass except of the second and the third step which are unified in the second pass.
If you are looking for a more detailed description then please read the tech guide.
The channel configuration checks the parameters of the incoming connection to detect misconfigured apaches and obsolete clients. The values in the configuration are regular expressions except of the type. The type defines the type of the environmentvariables which should be tested. Today we support only mod_ssl.
If you use encrypted connection then you must use
ssl
as protocol. If you need an unencrypted
connection like on the CA for the interfaces ca
and ca_node
then you must use
http
as protocol if you use mod_ssl. If you use
an Apache without mod_ssl then you must use .*
to match all incoming protocols. Please remember to set all the
keylengths to 0 because otherwise the access control rejects all
incoming connections.
Example 4.1. channel configuration
<channel> <type>mod_ssl</type> <protocol>ssl</protocol> <source>192.168.0.1</source> <asymmetric_cipher>.*</asymmetric_cipher> <asymmetric_keylength>0</asymmetric_keylength> <symmetric_cipher>.*</symmetric_cipher> <symmetric_keylength>128</symmetric_keylength> </channel>
We implemented three different ways to login to OpenCA:
none
passwd
x509
The first possibility means that there is no login and everybody who pass the channel verification can use the interface. This possibility is nothing else than the deactivation of the access control.
This is not only an option for debugging and testing. You can also use this option if you want to use a RA interface on a server which allows only RA access from the local machine but not over a remote computer.
<login> <type>none</type> </login>
This method can be used to login via login and passphrase. OpenCA supports authentication based on 1.) an internal database, 2.) ldap authentication and 3.) based on calling an external program to perform the actual user authentication.
It is possible to customize the headline (default: "Login to OpenCA") and the prompt for the login id (default: "Login"). This can be done with the following two configuration parameters:
Example 4.2. Login screen configuration
<login> <loginheadline>This is my customized login screen</loginheadline> <loginprompt>my login prompt</loginprompt> ... <login>
This feature is especially usefull if used with the LDAP authentication (see below), where you can use any attribute, such as email address as login name.
For the 'internal database' method you can specify one or more users. Every user has a name, a role, an algorithm and a digest. The algorithm specify which kind of digest should be used to hash the passphrase. OpenCA supports three algorithms SHA1, MD5 and crypt.
Example 4.3. Login and Passphrase configuration
<login> <type>passwd</type> <database>internal</database> <passwd> <user> <name>root</name> <algorithm>sha1</algorithm> <digest>3Hbp8MAAbo+RngxRXGbbujmC94U</digest> <role>CA Operator</role> </user> <user>...</user> ... </passwd> </login>
Before somebody tries to crack this hash the passphrase is root and this is the default passphrase of OpenCA :)
We prepared a script to generate the digests. The name of the script is openca-digest and it will be installed during make install*. The program is used like "openca-digest (help|sha1|crypt|md5) string".
If you have an LDAP server with login data about your users it makes sense to use it for user authentication in OpenCA. You have to configure the LDAP data for access, being hostname and port of the LDAP server, its base DN (the name space of the server), a bind dn of an entry for OpenCA to authenticate itself, which has to have appropriate access rights configured in the LDAP server, and the respective password, and whether OpenCa shall use TLS encryption in its communication to the LDAP server and the location of the respective CA certificate. Since the configuration file contains a password, it should only be readable for the OpenCA process.
Example 4.4. LDAP Login configuration: LDAP data
<login> <type>passwd</type> <database>ldap</database> <ldapdata> <host>ldap.foo.com</host> <port>389</port> <base>dc=foo,dc=com</base> <binddn>cn=openca,ou=services,dc=foo,dc=com</host> <bindpw>secret</bindpw> <usetls>yes</usetls> <cacertpath>/opt/certs/</cacertpath> ... </ldapdata> </login>
Next you have to configure, which LDAP attribute contains the identifier or login name of the users, which could be the uid or even an email address. Some such attributes may contain additional constant strings prefixed to the actual value (e.g. the attribute proxyAddresses when used for emasil addresses can contain the prefix "SMTP": "SMTP:misterx@foo.bar". If this is so you can specify the prefix, so OpenCA finds the user even if she does not include the prefix in the login name.
Example 4.5. LDAP login configuration: search attribute
<login> <type>passwd</type> <database>ldap</database> <ldapdata> ... <searchattr>uid</searchattr> <searchvalueprefix></searchvalueprefix> ... </ldapdata> </login>
The LDAP authentication module supports two different authentication methods:
bind (the generic simple LDAP authentication mechanism using the password stored in attribute userPassword)
pwattr (using the password stored in a freely configurable attribute, see below)
You can use both methods in parallel, but then the module must know which method to use for which entries. This can be defined by values of a certain attribute, which can be defined in the configuration as ldapauthmethattr.
Then you must define which values of that attribute should lead to which authentication method. A good example would be to take the attribute objectClass as ldapauthmethattr and say if the entry contains the objectclass posixaccount to use the ldap bind method, if it contains objectClass externalUser to use pwattr. such mappings can be done with the constructs <ldapauthmethmapping> (see example below).
If none of the conditions configured here are fulfilled by an entry, a default mechanism has to be used, which has to be configured (see example below).
For the pwattr method you need to specify which attribute contains the passwords to use. The values in that attribute can and should be stored as hash values. If so, the module needs to know which hashing algorithm was used. supported are: sha1, md5, crypt and none (=clear text).
Example 4.6. LDAP login configuration: authentication mechanisms
<login> <type>passwd</type> <database>ldap</database> <ldapdata> ... <ldapauthmethattr>objectclass</authmethattr> <ldapauthmethmapping> <ldapauthmethattrvalue>posixaccount</ldapauthmethattrvalue> <ldapauthmeth>bind</ldapauthmeth> </ldapauthmethmappingt> <ldapauthmethmapping> <ldapauthmethattrvalue>externalUser</ldapauthmethattrvalue> <ldapauthmeth>pwattr</ldapauthmeth> </ldapauthmethmappingt> <ldapdefaultauthmeth>bind</ldapdefaultauthmeth> <ldappwattr>mypasswordattribute</ldappwattr> <ldappwattrhash>sha1</ldappwattrhash> </ldapdata> </login>
The LDAP Login module also provides for role mapping, where certain values of a certain attribute map to certain OpenCA roles. First you have to specify which LDAP attribute contains the role mapping information. Then you can easily define the mappings (with constructs similiar to the above authmethmapping). This is configured within the <passwd> element:
Example 4.7. LDAP login configuration: role mapping
<login> <type>passwd</type> <database>ldap</database> <ldapdata> ... </ldapdata> <passwd> <roleattribute>memberOf</roleattribute> <rolemapping> <roleattributevalue>CN=OpenCA_RA,OU=UserGroups,dc=foo,dc=com</roleattributevalue> <rol>RA Operator</role> </rolemapping> </passwd> </login>
External authentication can be used to integrate authentication methods that are not natively supported by OpenCA. Essentially this method is a variant of the username/passwort authentiation.
External authentication requires an external program that receives the login credentials that were passed on the login screen via the environment. The external program then must verify if the username/password combination represents a valid login and return an appropriate exit code (0 for success).
The external program should take care of handling the username/password information properly, i. e. it should NOT write this information to files or pass these values to other programs on the command line. The latter is particularly important, as this might open security problems when processing untrusted user input (i. e. the login name or the password). If you must call external programs with this information, please take extra care to tidy the login information of illegal characters (such as quoting special characters). Specification of username or password as command arguments is not directory supported for exactly this reason. Use shell script wrappers that read environment variables instead.
Configuration is straightforward:
Example 4.8. External program authentication configuration
In this example configuration two environment variables USERNAME and PASSWORD are set for the external program/usr/local/bin/authdummy
.
The special strings '__USER__' and '__PASSWORD__' are
replaced with the actual user login information
within environment value definitions.
An arbitrary number of environment variables may
be defined in the configuration file.
<login> <type>passwd</type> <database>externalcommand</database> <setenv> <option> <name>USERNAME</name> <value>__USER__</value> </option> <option> <name>PASSWORD</name> <value>__PASSWD__</value> </option> </setenv> <command>/usr/local/bin/authdummy</command> </login>
As an example a very simple external authentication
program /usr/local/bin/authdummy
could (but should not) look like this:
#!/bin/bash if [ "$USERNAME" = "openca" -a "$PASSWORD" = "rocks" ] ; then exit 0 fi exit 1
Again, ALWAYS be sure to check the user input when processing the login data as arguments to external programs, it is easy to open gaping security holes here!
This is perhaps the most advanced method to login. The user must sign it's assigned session ID and the access control verifies the signature. The login name is the serial of the certificate because it is the only unique item in a certificate. The configuration is really simple. You have to set the position of the CA chain and that's all.
Example 4.9. Authentication with certificates
<login <type>x509</type> <chain>OPENCADIR/var/crypto/chain</chain> </login>
The filtering of the users is not the job of the login because the login has only to identify the user. The filtering has to be implemented in the ACLs. Therefore we cannot recommend the x509 (smartcard) authentication until now.
The session management is today really simple to configure because we only support one method to manage a session and the only real variable is the lifetime of a cookie after the last action. The lifetime must be value which will be accepted by CGI::Session. The directory contains the management informations for the cookies (like the name of the user).
Example 4.10. Session configuration
<session> <type>cookie</type> <directory>@var_prefix@/session/cookie</directory> <lifetime>1000</lifetime> <session>
The basic configuration of OpenCA's access control list is placed in OPENCADIR/etc/access_control/*.xml. The relevant datastructure is like follows:
Example 4.11. Basic ACL configuration
<acl_config> <acl>yes</acl> <list>@etc_prefix@/rbac/acl.xml</list> <command_dir>@etc_prefix@/rbac/cmds</command_dir> <module_id>0</module_id> <ca_cert>@var_prefix@/crypto/cacerts/cacert.pem</ca_cert> <map_role>no</map_role> <map_operation>no</map_operation> </acl_config>
Following you can find the meanings of the elements:
enable (yes) or disable (no) the access control list - please notice that a deactivated ACL means that every user has full access to ALL OpenCA functions
defines the place of the ACL
specify the directory which contains the configuration files of the scripts
This is the id of the interface. This id is unique for every interface.
The CA certificate in PEM format is used to verify signatures. You can use here another certificate than of the CA which you control with this access control. This is useful if you have a user CA and a server CA. You can login into the interface of the server CA with a certificate from the user CA.
enable (yes) or disable (no) the mapping from certificates to roles if the certificate of the user is known. If the role is defined during passphrase based login then this option causes the use of the specified role and not the login name.
enable (yes) or disable (no) the mapping from the names of the scripts to the supplied operation
You should check the access control list itself very carefully after you initialized the basic configuration of the interface. The real access control list is embedded into the access control area and has the following format:
Example 4.12. Permission for serverInfo
<acl> <permission> <module>0<</module> <role>root<</root> <operation>serverInfo</operation> <owner></owner> </permission> <permission>...</permission> ... </acl>
The meanings of the elements are the following ones:
every access right is defined in a permission element
this is the module id of the used web interface.
Every installed gateway of OpenCA is a module in the
terms of RBAC. If you install the RA then there is a new
module with the id 1
.
is the name of the user if map_role
is
no
and the role
of the certificate if map_role
is
yes
.
The roles are part of every role based system. OpenCA
defines a set of default roles which you can simply
extend by other roles which you need.
Every certificate will be assigned a role if it is issued on the CA. The role of a certificate service request is the role which the requests asks for. The role of a certificate revocation request is the role of the certificate to which the CRR belongs. The CA-certificate(s) and the CRLs have no explicit role because they have automatically the “superrole”. If there is an action where the user is not identified by a certificate then the role which is used is automatically the empty role. This is sometimes necessary for example if you want to control your public gateway by RBAC.
is the name of the script if
map_operation
is
no
and the operation of the
script if map_role
is
yes
.
is the role of the object which will be automatically detected by the configuration of the script
All options can be used with regular expressions but of course the parameters are case sensitive. An ACL which allows anything and only requires a valid login looks like this:
Example 4.13. Allow all
<acl> <permission> <module>.*</module> <role>.*</role> <operation>.*</operation> <owner>.*</owner> </permission> </acl>
Every command or script has it's own configuration file
which contains the name of the command (this
is actually a protection against the renaming of files), the name of the
operation for which it is used, the way how to find the affected object and
the name of the variable which contains the data which is necessary to
determine the object by the specified way.
Today there are six OWNER_METHOD
s:
CERTIFICATE_SERIAL
This method is used if an operation affects a certificate and the role should be detected by the serial of the certificate.
REQUEST_SERIAL
This method is used if an operation affects a CSR and the role should be detected by the serial of the CSR.
CRR_SERIAL
This method is used if an operation affects a CRR and the role should be detected by the serial of the CRR. The CRR will be loaded and the certificate which should be revoked will be loaded and the role of the certificate is used.
CGI
The use of this method is not recommended because the role is not protected by any cryptographic mechanisms.
ANY
The operator must have the right to perform this operation for every role.
<empty>
This method is used to signal that an object is handled which affects the CA directly (e.g. CA-certificate, CRL). The operator needs access to the “superrole”.
OpenCA has a concept which abstracts every key. Every key is a token and
be putted into a slot. This mean that the software can ask the
cryptolayer for
a token ca
and the cryptolayer checks it's
configuration for a token called ca
.
Now we will explain the token configuration because this is the most interesting thing for an administrator. The basic schema is the following one:
<openca> <token_config> <default_token>CA</default_token> <token>...</token> ... </token_config> </openca>
Every token configuration includes some common parts which will be described here. The names of the options are tag names!
which will be used by the software for the token. There are
four names today - CA
,
BP
(batch processors),
KEYBACKUP
and
LOG
.
This defines the type of the token. Today there are three
types (OpenSSL
,
Empty
and
LunaCA3
) but we can add modules for
every token you need if the token is supported by OpenSSL.
We add a module per token to be able to handle the different
details and proprietary tools to manage the tokens.
The mode describes how the token should be used. OpenCA
knows three modes standby
,
session
and
daemon
. standby
means that you must login to the token for every action,
session
activates the token for the
whole user session and daemon
runs the
token in a daemon mode which means that the token must be
stopped explicitly. Please read the docs about the different
token types because not every token supports every mode.
If the token login requires a password which the user have
to enter at the webpage then you should specify a sheet
where the user can do it. Usually there are prepared sheets
already installed at
OPENCADIR/lib/servers/server_name/sheets/token_login.html
.
So you have only to specify the default sheet.
We will describe now the configuration of different token types.
OpenSSL only support the operational mode
standby
. Additionaly it support several other
dynamic options which are required to work properly.
This is the path to binary of OpenSSL. It is called SHELL because it is the cryptoshell which we use. Simply run the binary without any options and you know what we mean.
filename of the file with the private key
the number of the components of the passphrase. If you
have two groups of administrators and every group has
only one part of the passphrase then you can enter
2
and OpenCA will display two
seperate input fields for the different parts.
path to the PEM formatted certificate
path to the DER formatted certificate
path to the TXT formatted certificate - this is only important for the CA token.
directory which includes the certification path - this is not only important for the CA because the chain is sometimes included into signatures.
This token is only a dummy if the key is not a really seperate key. Often the administrators simply use symlinks to the CA keys and certs for the keybackup etc.. An empty token is redirect to the default token which is usually the CA token.
This token accepts all the options like OpenSSL tokens except of
PASSWD_PARTS
because Chrysalis-ITS Luna CA3 uses
hardware (PIN pad) for login to prevent electronic reconnaisance
actions. Luna CA3 supports all operational modes
(standby
, session
and
daemon
. This module requires some additional
options:
Luna CA3 comes with a utility to manage things like login and keygeneration. You have to enter the complete path here.
This is an ID for the slot of the token.
This is the application ID to avoid conflicts with other
application. Please remeber that the
APPID
must be lesser than the
SLOT
.
OpenCA uses this file to detect that the token is already activated if the token runs in daemon mode. There is no way to find out this fact directly via a tool from Chrysalis-ITS.
This module provides basic support for nCipher HSM tokens and was tested with the following configuration:
SuSE Linux 8.1 and SLES 8
nCipher software
hwsp: 0.0.24cam37, 0.0.34cam7
hwcrhk: 1.9.0cam27, 1.9.7cam24
ctls: 0.0.24cam12, 0.0.32cam27
nCipher Modules verified (others may work)
nShield F3 SCSI module (nC3031S)
nShield F2 PCI module (nC1002P/nC3022P)
nShield F2 SCSI module (nC4022W)
The current status is considered to be stable.
Features:
Uses nCipher hwcrhk application and 'with-nfast' wrapper to perform nShield private key operations. Works with OpenSSL static and dynamic engine support (automatic detection from configuration).
Detects a number of problems and error conditions:
Middleware not running/operational (HSM online test)
Private key not available (key online test)
Current limitations:
Key generation is not supported (and will probably never be)
PIN entry is not supported
The module was derived from OpenCA::Token::OpenSSL and other existing token implementations.
All external program calls are subject to a hard timeout that is initially set to 15 seconds. This value can be configured by setting CHECKCMDTIMEOUT to the desired value in the token configuration file.
If an external program does not terminate within a certain timeout, it is killed by SIGALRM and the method fails with a timeout error (7151015). This was introduced in order to handle hanging nCipher utility processes after e. g. switching of an external SCSI HSM.
genKey() - This method is administratively disabled and always returns the error code 7154001.
online() - The online test performs the following tests:
Run 'enquiry' and check return code and program output
verify that the nCipher server daemon is operational
verify that at least one nCipher module is operational
The method returns true without calling external checks if the module accessibility was checked within the last 60 seconds. This value can be configured by setting the variable ONLINECHECKGRACEPERIOD to the desired value in the token config file.
keyOnline() - The following tests are performed in order to determine the key online status:Run 'nfkminfo' with the WRAPPER command (usually 'with-nfast -M') and check return code and output.
verify Security World status (must be initialized and usable)
verify that at least one object is preloaded
get object hash for private key to be used and verify it against the list of preloaded objects
The module requires a properly set up nCipher "Security World" including a private key to operate. Key generation is not supported by the module (as it really does not make much sense for a HSM based CA).
See Section 2.4.6, “Example for the setup” for an example.
In order to use the HSM protected key, log in into the CA computer and run /opt/nfast/bin/with-nfast pause in a shell.
You will be prompted to insert as many Operator Cards and input their corresponding PINs as required by the Operator Card Set Quorum.
As long as 'with-nfast' is not interrupted and the last Operator Card is not removed, any application with the proper Unix permissions may use the keys protected by the Operator Card Set.
To log out from the module it is recommended to terminate the with-nfast program (pulling the SmartCard works, too, but is not recommended, because it leaves the program waiting).
A simple login mechanism can be implemented using a special HSM login user. The user must have write access to the file /opt/nfast/kmdata/preload/default. For example create a user 'hsmlogin' with primary group 'kmdata'.
Change ownership and permissions of directory /opt/nfast/kmdata/preload so that 'hsmlogin' can enter ("execute") this directory.
Change ownership and permissions of the file /opt/nfast/kmdata/preload/default so that 'hsmlogin' can write to this file and that the www user running the ca CGI script can read it.
Create a login wrapper shell script /usr/local/bin/hsmlogin and assign this as login shell to the 'hsmlogin' user:
#!bin/sh exec /opt/nfast/bin/with-nfast pause
Login as hsmlogin in order to login into the HSM, use Ctrl-C to logout.
The key name should be the same as reported by /opt/nfast/bin/nfkminfo -k. This configuration reflects the use of OpenSSL static engine support. In order to use dynamic engine (e. g. when using OpenSSL 0.9.8), it is necessary to specify the location of the dynamic engine nCipher shared library in the token configuration:
<option> <name>PRE_ENGINE</name> <value>SO_PATH:/usr/local/openssl-snap/lib/engines/libncipher.so</value> </option>
Be sure to specify the correct location of the dynamic engine lib for the OpenSSL binary specified with the SHELL setting.
If at least one PRE_ENGINE setting is specified, the nCipher token module will automatically switch to dynamic engine semantics when talking to the OpenSSL backend. Definition of SO_PATH is the only mandatory setting in this case, and the following default PRE_ENGINE settings are automatically added (unless explicitly specified in the token configuration file):
ID:chil
LIST_ADD:1
LOAD
THREAD_LOCKING:1
Example for creation of a Security World, Operator Cards and CA key.
Assumptions:
example uses a 2 of 3 quorum ("2/3") for Adminstrator Cards and Operator Cards
the Operator Card set protecting the CA key will be named "RootCA"
the CA key will be named "rootkey" in this process
Sample Root Key ceremony:
initialize security world
switch HSM to 'initialize' mode
reset the module: /opt/nfast/bin/nopclearfail -c -m 1
/opt/nfast/bin/newworld --initialize --acs-quorum 2/3
switch HSM to 'operational' mode
reset the module: /opt/nfast/bin/nopclearfail -c -m 1
verify that the security world has been created
/opt/nfast/bin/nfkminfo
optionally: erase cards
/opt/nfast/bin/bulkerase -m 1 -s 0
initialize Root CA operator card set
/opt/nfast/bin/createocs --name=RootCA --ocs-quorum=2/3 -m 1 -s 0
verify that the operator card set has been created
/opt/nfast/bin/nfkminfo -c
create Root CA key
/opt/nfast/bin/generatekey2 --cardset=RootCA hwcrhk (example values: token, 0, RootCA, yes, RSA, 2048, 0, "", rootkey, no)
verify that the root key has been created
/opt/nfast/bin/nfkminfo -k
This token accepts all the options like OpenSSL tokens but
soem options has another meaning. The token only operates
in mode standby
. Other modes are not
supported by OpenSC today. The available options have the
following meanings (please notice that we only report
OpenSSL options if there meaning differ from the original
OpenSSL module):
The KEY is a combination of the slot and the ID of the
key on the card. This can differ for other PKCS#11
modules than OpenSC. If you use this class with
another PKCS#11 than the one from OpenSC then please
read the docs of the vendor. The syntax is
slot_[0-9]+-id_[0-9]+
. A
typical example is slot_0-id_45
.
The value should be always pkcs11
here. This defines that the OpenSSL dynamic engine pkcs11
should be used. The module is only tested with this
module.
These options are used to configure the PKCS#11 engine
which is loaded to OpenSSL.
SO_PATH
is the path of the used
OpenSSL engine. ID
is the used
OpenSSL engine ID. MODULE_PATH
is the path to the used PKCS#11 driver. Please notice
that we put the complete and partially internal
configuration into this parameter. The reason is that we
don't want to reduce the flexibility because we don't
know the future direction of the OpenSSL engine
interface and the OpenSC PKCS#11 engine for OpenSSL.
This is the name of the OpenSC carddriver. This option is only necessary if you want to create the key with OpenCA.
This is the number of the OpenSC carddriver. This option is only necessary if you want to create the key with OpenCA. (It is identical with the slotnumber.)
The path to the command pkcs15-init.
The path to the command pkcs15-tool.
The path to the command opensc-tool.
token.xml
with a freely
choosable name.
You must care about three configurationfiles and -directories
etc/openssl/openssl.cnf
,
etc/openssl/openssl
and
etc/openssl/extfiles
. The first file contains the
configuration for the CA. This means the file is used for the generation
of the initial CA-CSR, the selfsigned certificate (if you setup a Root
CA) and the CRLs. The file is configured fullautomatically during the
installation but if you are setting up a serious CA then you should check
this file too. The directory etc/openssl/openssl
contains the configuration for the different roles except of the
extensions. The relevant things which you must compare with your policy
are the lifetime of the certificate and the algorithm which is used to
sign the certs. The dircetory etc/openssl/extfiles
contains the definitions of the extension. Please check these files
carefully.
[[RFC3280] The authority key identifier extension provides a means of identifying the public key corresponding to the private key used to sign the certificate. This extension is used where an issuer has multiple signing keys (either due to multiple concurrent key pairs or due to changeover). ]
The authority key identifier is used for path construction. Only if you create a self-signed root CA certificate then you only need the subject key identifier. You can use the subject key identifier of the CA certificate or/and the issuer's certificate serial and issuer name. It is recommended to use both. It is important to understand that the name is the name of the issuer of the CA certificate. If you have a root CA, a sub CA and a user certificate then the name in the authority key identifier is the subject of the root CA's certificate.
The value should look like this:
Example 4.16. OpenSSL configuration - Authority Key Identifier
authorityKeyIdentifier=keyid,issuer:always
Never mark this extension as critical.
The different names of HTTPS servers are one of the most problematic things in the todays world. Like for many other cryptographic issues in the web there is a standard for servercertificates - RFC 2818 “HTTP over TLS”.
The standard defines that you have to check the subject alternative name for an appropriate entry (DNS or IP see RFC 2459). If this search fails then check the common name in the distinguished name of the certificate.
If you use Microsofts Internet Explorer then you have no problems. The IE is full standard compliant. The problem is Netscape. They use the common name like a regular expression in Unix. The common name can be in the format “(server1|server2).my_domain.org”. The clever ones would argue now that we must simply set the subject alternative name like defined by RFC 2818 and set a normal common name because the subject alternative is checked first. This is a nice idea but Netscape ignores the subject alternative name if it checks the name of the server versus the content of the certificate.
The solution is a mix between RFC 2818 and Netscape behaviour. You must set the common name in the distinguished name like Netscape defines it. After this you must set all DNS-names and the IPs of the server in the subject alternative name. If you do this then all standard compliant browsers will evaluate the subject alternative name first and will ignore the common name in the distinguished name. So the certificate is standard compliant but supports the cruel behaviour of Netscape too.
Before you think this is a perfect solution then please think about aliases. My personal record are 20 different names for one computer which I have to code in a common name. If you think that it is easy then please remember that a common name can only be 64 characters long.
Mailservers usually include SMTP daemons. SMTP servers act as server and client because they work in a hierarchy. Some server softwares like sendmail require that the SMTP server identifies itself as a SMTP client if it contacts another SMTP server. Usually you only want to issue one certificate per server and not one certificate per service and therefore you have to set the extensions for SSL Client and SSL Server like recommended by OpenSSL (please see "man x509" after you installed OpenSSL). If you use sendmail then you can create a server certificate for the SMTP server and an additional client certificate. Sendmail supports two certificates per server or better per daemon.
SSL Client requires the following extensions:
Example 4.17. Minimal SSL client extensions
keyUsage = digitalSignature extendedKeyUsage = clientAuth nsCertType = client
SSL Server requires the following extensions:
Example 4.18. Minimal SSL server extensions
keyUsage = digitalSignature, keyEncipherment extendedKeyUsage = serverAuth, msSGC, nsSGC nsCertType = server
msSGC and nsSGC mean Server Gated Crypto from Microsoft and Netscape.
If you want to configure only one certificate per SMTP server then you have to issue certificates which looks like a christmas tree. They have to include all extensions for clients and servers. A configuration can look like this:
Example 4.19. Minimal SMTP extensions for a single certificate
keyUsage = digitalSignature, keyEncipherment extendedKeyUsage = clientAuth, serverAuth, msSGC, nsSGC nsCertType = client, server
If you want to use OpenCA with F-Secure VPN+ then you should bear in
mind that this software can only download a CRL via http or ldap.
They don't support https and ldaps. This is important if you
configure your CRLDistributionPoints in
OPENCADIR/etc/openssl/extfiles/*.ext
. You can
easily fix this problem by using LDAP for CRL-distribution.
Certificates for VPN+ Gateways and Machine certificates should include the DNS name and IP address in the subject alternative name. The certificates for the persons to authenticate them can be contain anything you want. It must only be valid client certificates.
First we describe the concept of additional attributes and then we describe the two general types of requests - external prepared PKCS#10 requests and during the interaction generated requests.
Usually the first question is about what does an additional attribute be? Additional attributes where introduced to store extra informations in a request without publishing these informations. Big organizations or trustcenters need a lot of information and only a minimal subset should be public. If you are a postmaster or a webmaster of a university then it is a good idea to put the general emailaddresses into the certificates but it is not really optimal to store the telephonenumber in the certificate. Nevertheless it makes sense for the trustcenter stuff to have the phonenumber in a case of emergency.
After you know what a nice feature these attributes are you want to customize them? No problem. A simple example should help:
Example 4.20. Additional attributes configuration
ADDITIONAL_REQUEST_ATTRIBUTES "requestercn" "email" "department" "telephone" ADDITIONAL_ATTRIBUTES_DISPLAY_VALUE "Name (first and Last name)" "Email" "Department" "Telephone" ADDITIONAL_REQUEST_ATTRIBUTES_STRING_TYPE "LATIN1_LETTERS" "EMAIL" "LATIN1_LETTERS" "LATIN1_LETTERS"
The three options have the following meanings:
Table 4.1. Additional attributes configuration
Parameter | Description |
---|---|
ADDITIONAL_REQUEST_ATTRIBUTES | These are the internal names for the attributes. |
ADDITIONAL_ATTRIBUTES_DISPLAY_VALUE | Here you have to define the displayed names for the attributes. This is useful for example if you have two public interfaces for example an english and a german one. |
ADDITIONAL_REQUEST_ATTRIBUTES_STRING_TYPE |
This helps the software with the errordetection. You can
use the following values:
|
The first certificates which were needed in the open source are were server certificates. The most systems which use such certificates are Apaches, OpenLDAPs, IMAPDs, POPDs, SMTPDs and S-Tunnel. Such systems generate the private key by itself - means the administrator generate a key by hand or via the software but there is no interaction with the trustcenter's software until the administrator created a request with the key or exported a request from the software.
If the administrator has the PKCS#10 request then he must bring the
request to the trustcenter and this is the job of command which's
configuration we describe here. The option
DN_TYPE_PKCS10_REQUIRED_ELEMENTS
define the
structure of the subject of the PKCS#10 request. The option
DN_TYPE_PKCS10_BASE
and the values of
DN_TYPE_PKCS10_BASE_*
define the suffix for the
certificates which will be accepted by this interface. The
restrictions were implemented to get some kind of useful subjects.
Example 4.21. PKCS#10 configuration
DN_TYPE_PKCS10_REQUIRED_ELEMENTS "CN" "OU" "O" "C" DN_TYPE_PKCS10_BASE "O" "C" DN_TYPE_PKCS10_BASE_1 "OpenCA" DN_TYPE_PKCS10_BASE_2 "it"
If you have no prepared PKCS#10 request then there is a second method in OpenCA. This method is used if the key and request generation happen during the interaction with the client. OpenCA support clientside keygeneration for Microsoft Internet Explorer, Netscape Communicator, Mozilla and Opera. If you have another browser then OpenCA uses it's serverside key and requestgeneration. So let's start with an example:
Example 4.22. Basic CSR configuration
Basic_CSR_Keysizes "512" "768" "1024" "2048" "4096" DN_TYPES "BASIC" DN_TYPE_BASIC_KEYGEN_MODE "SERVER" DN_TYPE_BASIC_KEYGEN_SHEET "@lib_prefix@/servers/@pub_prefix@/sheets/basic_csr_confirm_request.html" DN_TYPE_BASIC_BODY "Y" DN_TYPE_BASIC_BASE "O" "C" DN_TYPE_BASIC_ELEMENTS "emailAddress" "CN" "OU" DN_TYPE_BASIC_NAME "Basic User Request" DN_TYPE_BASIC_BASE_1 "@ca_organization@" DN_TYPE_BASIC_BASE_2 "@ca_country@" DN_TYPE_BASIC_ELEMENT_1 "E-Mail" DN_TYPE_BASIC_ELEMENT_1_MINIMUM_LENGTH 7 DN_TYPE_BASIC_ELEMENT_1_REQUIRED "YES" DN_TYPE_BASIC_ELEMENT_2 "Name" DN_TYPE_BASIC_ELEMENT_2_MINIMUM_LENGTH 3 DN_TYPE_BASIC_ELEMENT_2_REQUIRED "YES" DN_TYPE_BASIC_ELEMENT_3 "Certificate Request Group" DN_TYPE_BASIC_ELEMENT_3_SELECT "Internet" "Partners" "Employees" "Trustcenter" DN_TYPE_BASIC_ELEMENT_3_MINIMUM_LENGTH 8 DN_TYPE_BASIC_ELEMENT_3_REQUIRED "YES"
The first line defines the available keysize. The next variable
DN_TYPES
defines the available configurations of
basic_csr. The command
basic_csr is called via a link and the link must
contain an option CSR_TYPE
which defines the
configuration which is used for this CSR. If you don't set this
option then basic_csr starts it's
browserdetection.
The default type which is supported by OpenCA is
BASIC
. You
can simply add a type and set a correct link in the public gateway.
You can find an example on the public gateway by looking at the link
“Basic Request”.
The prefix of every definition is now
DN_TYPE_BASIC_
. The NAME
defines the displayed name (e.g. "Request for managers only"). The
BODY
defines the type of the request. If the value
is Y
or YES
then a
key and a request will be stored and if necessary generated. If the
value is N
or samp
then
only a header will be generated. This option is used to get the
necessary data from a user to initialize a smartcard on the
registration authority.
DN_TYPE_BASIC_KEYGEN_MODE
specifies the way how to
generate a key and request. The supported modes are
SERVER
, SPKAC
and
IE
. SPKAC
is used with
Opera, Mozilla and Netscape, IE
is used for
Microsoft Internet Explorer and SERVER
is used
for all other situations.
The BASE
is the part of the subject which is not
editable by the user who requests a certificate. The other
BASE_numbers
define the values of the elements
which are used for the not editable part of the subject.
The ELEMENTS
are the part of subject which can be
defined by the user. The defined attributes of the subject can be
configured more precisely by the options named
*_ELEMENT_number*
. They have the following meaning:
Table 4.2. Generic basic CSR configuration
Parameter | Description |
---|---|
*_ELEMENT_number
| These are the displayed names of the elements. The normal user don't know what is a CN or a commonName. The most users will be confused if they see two fields with the same name (e.g. OU). So you can give the attributes some names which match their semantic. |
*_ELEMENT_number_MINIMUM_LENGTH
|
This field defines what the minimum length of the value
is. If you don't know it then simply use
0 .
|
*_ELEMENT_number_REQUIRED
|
Usually the user has to fill all fields but sometimes it
is a good idea to have some optional fields. If you have
such an optional field then please set the value to
something different than
“YES” . If a value is
entered by the user then the option
*_ELEMENT_number_MINIMUM_LENGTH still
will be checked.
|
*_ELEMENT_number_SELECT
|
All fields are textfields by default. You can
specify *_ELEMENT_number_SELECT
followed by a list of values. OpenCA creates a
HTML-select from this definition.
|
*_ELEMENT_number_XML_FILE
| If you need some more options or you have an export from an ERP database then there is an additional method to create HTML-select fields. You can create a XML file which must contain a list. The filename you must specify here. |
*_ELEMENT_number_XML_PATH
| If you specified an XML file then you have to specify the XPath to this list too. The XPath for the following example is “basic_csr/basic/element_3/option”: |
OpenCA displays at every time DNs like defined by RFC 2253. There are five options which influence the subject during the issuing itself:
Table 4.3. Common stuff configuration
Parameter | Description |
---|---|
SET_REQUEST_SERIAL_IN_DN |
This options enforce the inclusion of the request's
serial in the subject of the certificate. This is a
simple method to guarantee that the subject is unique.
True values are
Y , YES and
ON .
|
REQUEST_SERIAL_NAME | If the serial of the request will be included then this option defines which attribute is used for the serial. |
SET_CERTIFICATE_SERIAL_IN_DN |
This options enforce the inclusion of the certificate's
serial in the subject of the certificate. This is a
simple method to guarantee that the subject is unique.
This option is more recommended than
SET_REQUEST_SERIAL_IN_DN because the
value is tranparent. True values are
Y , YES and
ON .
|
CERTIFICATE_SERIAL_NAME | If the serial of the certificate will be included then this option defines which attribute is used for the serial. |
DN_WITHOUT_EMAIL |
This option is used to enforce recommendations of S/MIME
v3. If you don't want to include the emailaddress in the
subject then you can use this option. OpenCA will remove
the emailaddress from the subject before it issues the
certificate. True values are again
Y , YES and
ON .
|
OpenCA uses by default the old “o=University,c=de” style. Several users like international companies, universities or other big organizations need the new dc style. Therefore we support the dc style too. It is necessary to change several files because the configuration of the subjects is highly integrated into the software. We will explain it with an example.
base dn or suffix: dc=university,dc=edu user dn: dc=mike tester,dc=university,dc=edu webserver dn:dc=www,dc=university,dc=edu ca dn:dc=CA,dc=university,dc=edu
There are five things which you have to check for the change to the dc style. The steps will be now described:
There are two things which must be changed in the configuration files of the servers.
The LDAP configuration must be adapted to the new dc-style. The
variables - which you must modify - are basedn
and ldaproot
. The basedn
is
the suffix of the LDAP server. The ldaproot
is
the dn of the user root to bind to the LDAP server. The
ldaproot
has not to be changed because it is
freely configurable by the administrator of the LDAP server.
basedn "dc=university,dc=edu" ldaproot "dc=manager,dc=univesity,dc=edu"
The configuration of the requests must be changed too because they are prepared for the old style. Please read the following example to get an overview of a dc-styled configuration. Please read the section about the CSR configuration to understand how the normal requests can be configured.
DN_TYPE_BASIC_BODY "YES" DN_TYPE_BASIC_KEYGEN_MODE "SERVER" DN_TYPE_BASIC_KEYGEN_SHEET "/usr/local/OpenCA/lib/servers/pub/sheets/basic_csr_confirm_request.html" DN_TYPE_BASIC_BASE "DC" "DC" DN_TYPE_BASIC_ELEMENTS "DC" DN_TYPE_BASIC_NAME "Basic User Request" DN_TYPE_BASIC_BASE_1 "University" DN_TYPE_BASIC_BASE_2 "edu" DN_TYPE_BASIC_ELEMENT_1 "Name"
Please check the installed or prepared files with the name
main.html
because several HTML files
display the suffix of all the DNs.
You can find these files in
lib/servers/ra/mails
. They are the default
templates for the mails which RA Operators can send to the
users. They include the suffix of the LDAP server. This suffix is
called Dir Root
. This suffix must be changed
according to the real suffix of your LDAP server.
You must modify the files
OPENCADIR/etc/openssl/openssl.cnf
and
OPENCADIR/etc/openssl/openssl/*.conf
. The
policy and req sections must be changed to support requests and
certificates with subjects in the dc-style. If you don't know how to
configure OpenSSL then please read the documentation of OpenSSL.
If you generate the initial request for the CA request then please ignore all the fields for the normal subjectstyle. Simply enter nothing in all field until the software displays the window which show you the complete subject. There you have to enter the complete subject of the CA request. The subject is in RFC 2259 format and all “DC” must be written in big letters because OpenSSL is case sensitive.
The most people want to use OpenCA for issuing certificates to users or they want to test a PKI without buying a commercial trustcenter only for testing. So they want create a request, approve the request and issue a certificate. The problem is that they forget to the edit the request and the subjct alternative name was not set.
OpenCA knows two switches in ca.conf
to set the
subject alternative name automatically. The switch
AUTOMATIC_SUBJECT_ALT_NAME
enables
the mechanism to set the subject alternative name automatically if it
was not set in the header of the request. The second switch
DEFAULT_SUBJECT_ALT_NAME
defines the type of the
default value. Actually we implemented only support for the
emailaddress. If you need
support for DNS name(s) or IP addresse(s) then contact us. We only don't
implement it because nobody need it until now.
If you edit the subject alternative name on the RA interface then you
see only four fields where you can enter parts of the alternative name.
If you fill all fields then you will get at next time you want to edit
the alternative name one additional field. If you know that you need
more or less fields then you can change the option
CSR_MAX_SUBJECT_ALT_NAMES
in the configuration files.
The option defines the number of the displayed fields by default. It is
NOT a hard limit.
OpenCA provides an LDAP interface for users to download certificates from a central repository. This interface can be utilised by browser address books and specialised LDAP clients.
Before the OpenCA Online components can write certificates and CRLs to the directory you must have an LDAP compliant directory installed and available to the online components (this can be on the same or different machine). One example of an appropriate directory is the OpenLDAP project.
A full description of the configuration of your LDAP directory is outside the scope of this document. Important points to note are:
Ensure that the following schemas are included (probably in the slapd.conf file):
core.schema
cosine.schema
inetorgperson.schema
openca.schema
Ensure the directory is started with the appropriate suffix (e.g. o=myorg,c=gb).
Ensure the rootdn is specified.
Ensure the root password is specified.
Three configuration files must be configured for the online
components to make use of the LDAP directory to store certificates;
OPENCADIR/etc/servers/node.conf
,
OPENCADIR/etc/servers/ldap.conf
and
OPENCADIR/etc/ldap.xml
.
Usually it is enough to set the
correct options in ldap.xml
or in
config.xml
.
The configuration is splitted into two parts - a OpenCA related
part and a LDAP related part. The OpenCA related part consists
only of four options - the LDAP activation, automatic LDAP
updates during imports and distinguished name manipulations for
CA objects like CA certificates and CRLs. These options can be
configured in the interface configurations in
OPENCADIR/etc/servers/
.
Supported options in interface configurations
If you set this option to "yes" then the LDAP code will be activated.
This option will be used by the node interface. If the value is yes then the LDAP server will be updated automatically during imports of certificates, CRRs or CRLs.
Some users want to store the CRL in a special node of the LDAP server which is not identical with the issuer of the CRL. This can be happen if the user specifies a special CRL Distribution Point (CDP) which differs from the subject of the CA certificate. Here you can specify this special distinguished name. Please remember that OpenCA is today not able to add this node automatically if it is not present.
Some users want to store the CA certificate in a not standard conform node which means that there is perhaps an already existent directory which conflicts with the PKI structure. Here they can add the distinguished name of this special node. This node can be automatically added by OpenCA.
The LDAP related part of the configuration can be found in
OPENCADIR/etc/ldap.xml
. This central
configuration file avoids double configurations which can
produce many errors and confusion because you are sure that
you changed it but you only did it for one interface. The
isolated configration allows better names for the
configuration options too.
This is the hostname of your LDAP server.
This is the port where your LDAP server listens.
This is the suffix (OpenLDAP terminology) of your LDAP server. You can add here several suffixes if your LDAP server supports this feature (e.g. OpenLDAP v2). Every suffix must be placed in a seperate dn tag. The suffix tag is the bracket for all those suffixes.
The bind DN of the user which OpenCA uses to add data to the server.
The passphrase for OpenCA's ldap account.
OpenCA supports LDAP v2 and v3. The default is v2 because all servers can support v2. Several new distributions especially of Linux deactivates the LDAP v2 support. So if your OpenCA LDAP code completely fails check first the protocol versions of OpenCA and your LDAP server.
Some other options like ldaptls and ldapsasl require LDAP v3. So be really careful which protocol you use. If your LDAP server supports protocol version 3 then please use it. It avoids a lot of trouble.
Use no or yes to deactivate or activate TLS. Please remember that this option only works with LDAP v3.
Use no or yes to deactivate or activate SASL. Please remember that this option only works with LDAP v3. We use CRAM-MD5 for passphrase hashing.
OpenCA supports the possibility to exclude roles from certificate publishing. This can be useful for security reason and be required by privacy laws. If you have such a special role simply add it to to this options.
As long as the option updateLDAPautomatic
is set to
yes
the online components will attempt to upload
certificates to the directory after an import. Before this can
happen the directory must be initialised and the appropriate
structure must be implemented. In this version of OpenCA this
initialization is done automatically.
The common situation in large directory projects is a big schema
and at the end a small request for certificate integration. This
is usually no problem if OpenCA has to create and add
certificates to an existent and filled LDAP server. OpenCA gets
a problem if you want that OpenCA initialize this LDAP server
and should create all missing nodes in the directory tree. If
this is the case then you must integrate your schema
specification into OpenCA's ldap configuration. You can find
this schema specification in ldpa.xml
. The
path to the schema specification is
openca/ldap/schema
.
First you have to understand the general design of OpenCA's LDAP schema support. We have a very pragmatical idea of directory trees because it is not possible to handle all full featured ideas. We simply use the least significant attribute to select an appropriate schema definition for a distinguished name. Do you want to know what this mean? If you have a RFC 2253 conform distinguished name then you take the RDN on the left side. The used attribute type is used to detect the appropriate schema.
Before you can select a schema with the atribute type you must
know what you want to create for a node in the directory tree.
If the used distinguished name is the complete subject of a
certificate then you must select the schema specification from
the XML path openca/ldap/schema/certificate
.
Otherwise you you have to use
openca/ldap/schema/default
. If you need to
store a CA certificate then you must specify the schema in
openca/ldap/schema/ca
.
Every RDN entry in the described sections specifies the characteristics for one attributetype. This include things like required (must) or optional (may) attributes and used objectclasses (structural and auxiliary). If you have for example a new attributetype uid_special and a new class MY_CLASS then your RDN section for certificates with this attributetype as last one should looks like this.
OpenCA's schema definition is perhaps not so flexible as you need but we are open for new ideas. Be free to mail us your ideas. If they are not too proprietary then perhaps we can integrate them :)
SCEP is the successor of CEP the Certificate Enrollment Protocol. Both protocols were developed from Cisco. The idea is to have simple but secure protocol to enroll certificates and CRLs. Today many network components use SCEP to manage certificates and CRLs. Some of these components are Switches, Routers, Firewalls and VPN-Softwares.
OpenCA support SCEP via an own web interface. The interface is called
scep and you can install it via "make install-scep". After the
installation you have only to configure the file
OPENCADIR/etc/servers/scep.conf
or you edit
OPENCADIR/etc/config.xml
before you run
OPENCADIR/etc/configure_etc.sh. Please remember to
only filter via IP addresses because SCEP doesn't support any
authentication mechanisms. A SCEP client can connect the interface via
http://your_host/cgi-bin/scep/scep
.
The SCEP Server can accept the desired certificate role from new enrollment requests from the CertificateTemplate Extension. In order to make use of this feature, set the PKCS#10 attribute 1.3.6.1.4.1.311.20.2 to the desired certificate role name (Unicode) The Server will remove any non-alphanumeric characters and then try to match the requested role with the ones defined for the RA. If a match is found, this Role is selected for the new certificate request.
This file contains the followin parameters:
This is the PEM encoded private key of the SCEP interface. It has the same format like for mod_ssl.
This is the PEM encoded certificate of the SCEP interface. It has the same format like for mod_ssl.
This is the passphrase for the private key of the SCEP server. If you use a not encrypted private key (what is not recommended - then please set an empty string here. interface. It has the same format like for mod_ssl.
Policy definition, if set to "NO" the SCEP server will not accept requests for certificate DNs that do not exist yet. Set this value to "YES" to allow initial enrollment of new systems unconditionally. Experimental: If set to "VALIDSIGNATURE" then the server will only accept SCEP requests that have been signed with a certificate issued by the same CA. This is not strictly defined by the newer SCEP drafts but allows to add an additional authentication step. In this case no self-signed certificates are accepted in the SCEP message, the server will require a valid signature (but this may be of any certificate issued by the infrastructure).
Policy definition, if set to "YES" the SCEP server will allow renewal requests for existing certificates. A renewal request will not be accepted if two or more valid certificate with the same DN already exist. If set to "NO" the server will reject SCEP requests for DNs that already exist in at least one valid certificate.
String containing a comma separated concatenation of all RDNs to consider for matching the incoming request DN against existing valid certificate DNs. The reason for this setting is that some SCEP clients have problems generating request DNs that conform to the CA conventions. By specifying which RDNs to consider from the incoming request, it is possible to address these problems by ignoring unwanted RDNs. At least the CN should be included here. If your CNs are not unique across your name space, you should add other RDNs to create a unique reference, in most cases this would be "O", "OU" or "DC".
When initially enrolling a new system, this option specifies the initial certificate role to use for the new request. For renewals (at least one certificate with the same request DN already exists) the certificate role from the youngest existing predecessor is used for the new request.
When initially enrolling a new system, this option specifies the Registration Authority to use for the new request. For renewals (at least one certificate with the same request DN already exists) the RA setting from the youngest existing predecessor is used for the new request.
Newer SCEP drafts allow signing the SCEP request with a previously existing key. If a renewal request is received (this means that a valid certificate with the request DN must already exist in the database) the digital signature of this request is verified against the local CA. If the signature is valid and if the request DN is identical to the signer DN then the new request is automatically set to the 'APPROVED' status in the database. The client must explicitly support this feature by using the old certicate and private key. As the feature was introduced in later SCEP drafts, not many SCEP clients allow for automatic approval. Newer versions of sscep will support this feature, however. (Please note that the client system certificate may not have the correct key usage for digitally signing a request. The SCEP draft specifically allows the violation of the key usage, and hence OpenCA's SCEP server ignores errors about incorrect key usage when verifying the signature.)
This file contains the followin parameters:
This is the PEM encoded private key of the SCEP interface. It has the same format like for mod_ssl.
This is the PEM encoded certificate of the SCEP interface. It has the same format like for mod_ssl.
This is the passphrase for the private key of the SCEP server. If you use a not encrypted private key (what is not recommended - then please set an empty string here. interface. It has the same format like for mod_ssl.
The dataexchange of OpenCA is highly configurable. So first we have to describe some general concepts.
If you look at OpenCA from a database viewpoint then OpenCA is a tree of hierarchical organized databases. Every database is used by some web interfaces. So one node of the hierarchy consists of a database and some web interfaces. If we describe the dataexchange then we describe the dataexchange between nodes. This is the reason why we called the management interface node.
A node can exchange data with a node of a higher level of the hierarchy
or with several nodes which are on a lower level of the hierarchy. If
export data to a higher level of the hierarchy then we
UPLOAD
data and if we import data from such a node
then we DOWNLOAD
data. If import data from a lower
level then we RECEIVE
data and if export data to
such a node then we ENROLL
data.
If you exchange object in a security relevant area then you must define
which object with which state you want to exchange. Therefore you can
define in OpenCA which objects with which state you accept from which
direction. Also OpenCA allows only to overwrite existing objects if you
DOWNLOAD
CA-certificates, CRLs, CSRs and CRRs.
Status or object
injections are not accepted in all other situations. OpenCA includes some
default configurations to help you on the way to secure configuration.
The following example is for the import from a higher level of the hierarchy:
Example 4.27. Download configuration
DOWNLOAD_CA_CERTIFICATE_STATES VALID DOWNLOAD_CERTIFICATE_STATES VALID DOWNLOAD_CRL_STATES VALID DOWNLOAD_CRR_STATES ARCHIVED DELETED APPROVED DOWNLOAD_CSR_STATES ARCHIVED DELETED DOWNLOAD_MAIL_STATES CRINS DEFAULT
The export/import technology itself is another important aspect. You can
configure OpenCA to use different methods for the dataexchange with higher
and lower levels of the hierarchy. Therefore we implemented options to
prepare and cleanup IO operations. This is necessary if you have to start
special networkinterfaces or to mount devices. There are also options to
configure the EXPORT
and
IMPORT
of the directory with the
data and there is an option to TEST
the result.
EXPORT
, IMPORT
, START
and STOP
accept more then one argument.
Every argument will be seperately executed. The parameters
@__SRC__@
and @__DEST__@
include the directories with the data. The parameter
@__DEVICE__@
will be replaced with the value of
DEVICE
.
Example 4.28. Export configuration
EXPORT_IMPORT_UP_DEVICE "/dev/fd0" EXPORT_IMPORT_UP_START "" EXPORT_IMPORT_UP_STOP "" EXPORT_IMPORT_UP_EXPORT "/bin/tar -cvfp @__DEVICE__@ -C @__SRC__@" EXPORT_IMPORT_UP_IMPORT "/bin/tar -xvf @__DEVICE__@ -C @__DEST__@" EXPORT_IMPORT_UP_TEST "/bin/tar -tvf @__DEVICE__@"
The hierarchylevel LOCAL
is used for backups,
batchprocessors and such things. It is also possible to configure
OpenCA for the use with temporary private networks. Here is another
example for a CA:
Example 4.29. Local export configuration
EXPORT_IMPORT_DOWN_DEVICE "openca.tar" EXPORT_IMPORT_DOWN_START "/sbin/ifconfig eth0 up" EXPORT_IMPORT_DOWN_STOP "/sbin/ifconfig eth0 down" EXPORT_IMPORT_DOWN_EXPORT "/bin/tar -cvfp /usr/local/openca/var/tmp/@__DEVICE__@ -C @__SRC__@" "/usr/bin/scp /usr/local/openca/var/tmp/@__DEVICE__@ openca@ra.openca.org:/usr/local/OpenCA/var/tmp/" "rm /usr/local/openca/var/tmp/@__DEVICE__@" EXPORT_IMPORT_DOWN_IMPORT "/usr/bin/scp openca@ra.openca.org:/usr/local/OpenCA/var/tmp/@__DEVICE__@ /usr/local/openca/var/tmp/@__DEVICE__@" "/bin/tar -xvf /usr/local/openca/var/tmp/@__DEVICE__@ -C @__DEST__@" "rm /usr/local/openca/var/tmp/@__DEVICE__@" EXPORT_IMPORT_DOWN_TEST ""
The standard data exchange between RA and CA is done via
/dev/fd0
(floppy). If the CA and RA are really
big then it is recommended to change the data exchange location to a
simple file on the local system (for example
OPENCADIR/var/tmp/fd0
).
Please always remember that your apache must be able to access this
file. You can also do a chown and set the owner to apache's user.
You also have to edit the dataexchange sections of the .conf files in
OPENCADIR/etc/servers/
to point to the new file
(edit the lines EXPORT_IMPORT_*_DEVICE
).
It is recommended to erase the file after transfers, because the exchange can contain private keys of users.
If the CA and RA are located on different machines in a secure environment with perhaps an offline CA it can be difficult to do the dataexchange via simple files (the files have to be transferred between CA and RA, either via a diskette or ftp (not secure)). OpenCA offers the possibility to do the dataexchange automatically via scp.
The RA exports resp. imports from a file located on the system it is installed on (same as local datatransfer). Thus the RA configuration is the same as for the local dataexchange:
Create a new file for the export with the correct permissions like
OPENCAIDR/var/tmp/fd0
Edit the ra_node.conf file in
OPENCADIR/etc/servers/
to point to the new
file (edit the lines EXPORT_IMPORT_*_DEVICE
).
To use scp you must have a working openssh environment with an ssh client on the CA side and an ssh server on the RA side. On the CA side determine the apache home directory by looking at the etc/passwd file (in redhat 8 and 9 this is /var/ww). Do the following:
Now go back to the CA side to edit the files
ca.conf
and ca_node.conf
.
The ca_node.conf
file already contains a sample
scp datexchange section. Comment out the dataexchange section used for
local datexchange. Uncomment the scp part and edit it:
The EXPORT_IMPORT_DOWN_DEVICE
is the file used
for datexchange and must have the same name as the one defined
in ra_node.conf
on the RA side.
The EXPORT_IMPORT_DOWN_START
can have no value
or a value like "ifconfig eth0 up" in
which case the CA goes online for the dataexchange.
The EXPORT_IMPORT_DOWN_STOP
can have no value
or a value like "ifconfig eth0 down" in
which case the CA goes offline after the dataexchange.
In the EXPORT_IMPORT_DOWN_EXPORT
line we can
define the operations OpenCA has to perform when doing an export
(CA to RA transfer).
In the EXPORT_IMPORT_DOWN_IMPORT
line we can
define the operations OpenCA has to perform when doing an import
(RA to CA transfer).
In the EXPORT_IMPORT_DOWN_TEST
line we can define
the operations that have to be performed to test the transfer
(was it successful or not).
If you create a new node e.g. a second RA then you have to support this node with the dataexchange mechanism. Every interface of OpenCA must have a unique module ID. OpenCA manage the complete dataexchange with the ID of the node interface. The node interface knows which object of which datatype was already received by another node.
If you want to create a new node then you must create the corresponding
files in OPENCADIR/var/log
. You have simply to
create some files in the directories
OPENCADIR/var/log/enroll
and
OPENCADIR/var/log/download
depending on the
direction which you use for epxort. These directories
contain some files of the style $number_$datatype
.
$number is the
module ID of the node to which you want to export the data. The
datatype is from the exported objects.
If you created a new module ID (e.g. you setup another RA) then you
have simply to touch the file $number_$datatype
.
The new file is empty and so all objects will be exported.
This is only a short introduction to PostgreSQL. This section should not replace the reading of the PostgreSQL Administration Guide but after the lecture of this section you should be able to setup OpenCA very easily with a real datbase and not only with DBM files. This is important because PostgreSQL has transaction support which is essential for serious database applications.
If we talk about real world databases then it is a good idea to mention pgadmin III. The old versions pgadmin I and II only support windows. The third version supports Unix too. So there is now a GUI like for MySQL available which can be used to manage the database. There are people who think that GUIs are not necessary but they enhance the management of several things a lot because not everybody has the time and skill to optimize it's PostgreSQL database perfectly.
Another note about security, please never trust database security mechanisms fully. Use at every time a small IP-firewall in front of your database server. The default installation of OpenCA is a datbase server on localhost. If you want to install a database server on a different machines than the OpenCA components then always install these servers at minimum in a DMZ. Databases like PostgreSQL has today it's own strong security mechanisms but they have from time to time some big bugs. OS-based IP-filters like ipf or netfilter are usually more robust.
After the installation of PostgreSQL there is usually a use
postgres. Please use
su - postgres to use this identity. After this step
you should login into the database with psql.
This is possible because the database template1
is always present on new databaseservers.
The first real step is the creation of the user
openca with create user.
After this you logout and login again as the newly created user.
Now you create the database openca. After this the database itself is
ready for use but you should make the database a little bit more secure.
Sometimes the database is already secure then you have problems to issue
the psql commands. Please edit the pg_hba.conf
first to get the access permissions for the openca user.
You can restrict the database via PosgrSQL's own configuration. The
directory /etc/postgres
contains usually a file
pg_hba.conf
which controls the access to the
database. There should only be one entry which looks like follows:
This configuration restricts the access to the PostgreSQL database to the database openca via users from localhost. Nobodyelse can access the database now. Please remove all the other access rights - especially the lines which use IP-based trust settings. After the configuration you must restart PostgreSQL. The tables can be created with OpenCA's web interfaces.
Please ensure that the PostgreSQL server is connected to the network. Usually the TCP/IP socket is deactivated.
Sometimes the command of PostgreSQL changes. So please try to run the command \h in psql to see a list of the actual available commands.
PostgreSQL maintains three programs to dump a database - pg_dump, pg_dumpall and pg_dumplo. We describe pg_dump here because we don't like to backup all other databases and we don't like ot backup large objects only. The following options are used:
large object are included
create the database newly on recovery
export all data as insert commands to support on the fly database mogration
columnname based inserts to support rearrangement of column ordering
If we have different backup methods then we have different recovery procedures too. The main difference is that we can use the plain text export directly with psql to import the database. This is possible because this file only includes pure SQL statement. The PostgreSQL specific exports must be processed with pg_restore because they include support for BLOBs too. There is only one command for both PostgreSQL formats because pg_restore detects the used format automatically.
Until now nobody created a howto for MySQL.
If you use a version of MySQL prior 4.1 then you can see perhaps an
error message which reports a syntax error because of an unkown
internal variable NAMES
. You can ignore this error
message or better you should use a newer version of MySQL.
The background is the internationalization of OpenCA.
We support several
different character encodings. MySQL can only handle by default the
character encoding which is specified for the table during
configuration. The solution is the internal variable
NAMES
which was introduced in the SQL92 standard.
This variable can be used to set the actual character encoding.
MySQL supports this variable beginning with MySQL 4.1. It is
recommended to use at minimum 4.1.1 because of a rewrite of this
stuff in MySQL.
Oracle is supported by OpenCA via the standard Perl DBD Driver. However, there are some special considerations concerning system setup, configuration and security.
Oracle databases are accessed via the Perl DBD::Oracle module which uses Oracle OCI client libraries to access database services. Both must be installed on the system.
If the location of the Oracle OCI client shared library is incorrectly configured the OpenCA daemon will not start up properly.
As a consequence, the OpenCA application must be able
to locate the Oracle
shared library libclntsh.so
. Because
the client library is usually not installed in a standard
library path but rather below the Oracle home directory, the
explicit location of this library must usually be configured on
the system:
Explicitly set the directory location of
libclntsh.so
in environment variable
LD_LIBRARY_PATH for the OpenCA daemon. To do so you could
for example write a wrapper script that calls
etc/openca_start or include the
environment setting in the openca_start
script itself.
Change the dynamic linker configuration by adding the required directory location to the config file (/etc/ld.so.conf on a Linux system). You will have to ensure that the dynamic linker configuration is updated (Linux: run the ldconfig command).
The parameters db_host and db_port are completely ignored for Oracle databases, make sure to leave them empty in order to avoid confusion.
In config.xml
set the db_type to 'Oracle'.
The db_name, db_user, db_passwd and db_namespace variables may or
may not be set according to your configuration (see below).
The Oracle client library requires some environment variables that
must be set. As a minimum, the ORACLE_HOME variable must be set.
This can be done via the etc/database/DBI.conf
configuration file or by sourcing the oraenv
file that should have been set up for your Oracle instance by your
Oracle DBA.
Depending on your setup it may be useful to source the
ora_env
file in the OpenCA rc start script
like in the following example. In this case, usually no
environment variables need to be set up in the
DBI.conf
file.
Internal authentication is the classic authentication method and is also the easiest to configure. It is similiar to the methods used for the other database types and requires the explicit configuration of a database user, a password and a database name. The disadvantage, however, is the need to configure the database password in the configuration file which is not always desired. If this is an issue, consider the external authentication method discussed below.
The db_user and db_passwd parameters must be set to the correct username and password values for the OpenCA database. These are identical to the parameters specified in the CREATE USER user IDENTIFIED BY "password"; command that is used to create the user in the Oracle database.
The db_name parameter must be set to the database name configured
in the TNSNAMES.ORA
configuration file. As
it is possible to specify aliases in this file, this may also
point to an alias name of the database. In order to verify if your
db_name is set correctly run the command tnsping
with the db_name parameter as its only argument. The command must
be run with the same environment variables that are set for the
OpenCA daemon.
If the tnsping command above fails then the database setup is incorrect and must be fixed. Debugging this error is beyond the scope of this document, as there are lots of possible error causes.
If privilege separation should be used (see below) then the db_namespace parameter must be set to the name of the database user owning the OpenCA schema.
Internal authentication is easy to configure but has the significant drawback of requiring a cleartext password for the database user in the configuration file.
External authentiation grants a Unix user access to the database without the need of a password.
This authentication type usually only works for databases running on the local machine. Although it is possible to configure external authentication for remote machines this is not recommended due to security reasons.
In order to set up an Oracle user for external authentication for the Unix user 'wwwdata' use the following Oracle command (the string 'OPS$' must be prefixed to the Unix user name): CREATE USER OPS$WWWDATA IDENTIFIED EXTERNALLY; After the required permissions (e. g. CREATE SESSION) are granted to the user you can test this by trying to access the database using sqlplus / as the user wwwdata. Make sure that the environment (in particular the ORACLE_SID environment variable) is set up properly for this user. The database monitor should now start without a password prompt.
To make this work with your OpenCA setup you must use the
Unix user name that is configured as httpd_user in your
etc/openca_start
startup file.
Assuming the 'OPS$...' user has been created and the database schema is already set up properly, the database configuration for OpenCA must be completed.
DBM file are much easier to handle than real SQL databases. If you want to use these database then you must only ensure that the directory which should contain the database files is fully writeable, readable and accessible by the webserver user. This will be handled by OpenCA's installation routine automatically. Sometimes in the past the users choose the wrong webserver user. The result is a message in the logs that the function configError doesn't exist. This happens because the OpenCA script cannot load the library files. The new versions of OpenCA (0.9.1.4+) display correct errormessages in this case. Extra actions by the installing administrator are not necessary.
Please never forget that DBM files don't support transactions. If you implement a real world PKI then it is strongly recommend to use SQL databases to have a consistent state of the PKI even in the case of a system crash. DBM has also problems with multi user access for example on web servers with high loads.
The following example shows how to configure sendmail to send mails via a mailserver which requires a login. This basic authentication is today very common to protect the mailservices especially of big providers against spammers. The example was developed and tested with Microsoft Exchange but it should work with any other mailserver (e.g. sendmail or postfix).
Table of Contents
This is the ultimate user guide for OpenCA. Here we want to answer all the questions which user can have if they use OpenCA.
Table of Contents
This section describes the public interface to the OpenCA PKI. From these screens a user can view current certificate lists, manage certificates and download CA and revocation certificates.
There are a set of "tabs" along the top of the screen, each is described below.
This section describes the current versions of the OpenCA modules. There is only one sub menu item, Logout.
This section describes the CA related utilities a user can access. Each heading below relates to a link under the CA Infos tab.
By hitting this link the user is presented with a page titled "Download and Install CA Certificates". This page contains links to CA certificates in various formats.
In order for the user to "trust" certificates generated through OpenCA they must have the Certificate Authority root certificate installed. This page provides an easy mechanism for them to do that. Most users will just click the "CA-certificate in format CRT" and follow the instructions presented to them by their environment (e.g. In IE they have the option to "Open" the file and then "Install Certificate").
Apache Web server administrators would use the link "CA-certificate in format PEM" to download the certificate in the appropriate format for inclusion in the Apache configuration files.
By hitting this link the user is presented with a screen entitled "Download and Install CRLs". This page contains links to certificate revocation lists in various formats.
Many certificate aware clients (like Microsoft Outlook and Netscape Navigator) make use of certificate revocation lists to ensure that certificates are still valid and have not been revoked.
Three links are provided each containing the current CRL in a different format depending on the client. Normal client users would download the CRL in DER format for inclusion in their browser. Web server administrators would use the PEM format. The text format is downloaded as a human readable file.
This section allows the user to manage their certificates. It allows certificate request, retrieval, testing and revocation.
This link presents the user with a page offering a number of certificate request methods. There are subtile differences between methods which are described below. Each one of the links will take the user to a form. The user will fill in the form and submit the data. The data submitted will be used to create a certificate signing request (CSR) which will go to the certificate Authority to sign and return as a certificate.
The form data has the following fields:
E-Mail: The email address associated with the certificate
Name: The name of the user
Certificate Request Group: This is usually the department or sub group the user belongs to
Alternative email: Another email to appear in the certificate
DNS name: The DNS name if the certificate is a web server cert
Name: The real name of the user
Email: The users email address
Department: The user's department
Telephone: The users telephone number
Level of Assurance: This is the type of physical authentication the user is to receive
Role: The certificate role within the hierarchy, this is usually "User" for most normal users
Registration Authority: This is usually the physical location at which the user is to be identified (e.g. Personnel)
PIN: A password used to verify the CSR
Key Size: The size of the key used in the CSR (Usually set to 1024
After submitting the form the next set of screens the user sees will depend on the client being used and the type of request selected. After the CSR has been generated and submitted the user will be issued with a Certificate Request ID. This takes the form of an integer number. It is important that the user notes this number down as it is required when retrieving their certificate.
Once the user has requested their certificate the Certificate Authority will process the certificate request. This may involve a face to face identification of the user at the Trust Center. When the certificate has been created the user will be informed by email. This email will also include a Certificate Revocation Number (CRIN), this number should be kept in a safe place as it will be required if the user to needs to revoke their own certificate in the future.
By pressing this link, OpenCA will try to determine what browser the user is using to request their certificate. Once this has been established the CSR form is presented to the user. The CSR (along with the associated private and public keys) will be generated by the user's browser and submitted to the Certificate Authority.
This link leads to a server side key and CSR generation. A user would use this link if their browser did not support CSR generation, of if for some reason they wanted the Certificate Authority to generate the keys and CSR (e.g. For key backup on the server).
This link should be used if the client is a Netscape type browser (e.g. Navigator or Mozzilla). The CSR generated by the client will be of the type SPKAC.
This link should be used if the client is an Internet Explorer type browser. The CSR will be generated by the client.
This link is used to submit a web server certificate request. This is slightly different from a normal client certificate request as the CSR will have already been generated at the web server. There is a field used to upload the CSR, so the user must make sure that they have a CSR to upload before selecting this option.
This link is the same as the "Basic Request" in that the keys and CSR are not generated at the client. This request is used when the Certificate Authority is going to create the key pair and certificate on a hardware token. You only enter your data and the data is stored at the server but no cryptographic operations take place without operator interaction. In simple terms it is like an email with "Hello, I need a cert. Sincerly your Jon Doe".
This link provides the mechanism for a user to retrieve the requested certificate and install it in the browser.
The user is presented with a screen and a set of instructions. The most important being that the user must be using the same computer that was used to request the certificate. This is important because both IE and Netscape type browsers need to link the certificate back to the CSR and private keys, this can only be done if the computer that was used to generate the CSR is used to retrieve the certificate.
The user should enter their "Serial Number" in the space provided. The serial number can be:
The serial number of the certificate signed by the Certificate Authority
The serial number of the submitted request issued to the user at CSR submission time
This is the ID which i used for the batchprocessor. Usually this ID is an account but the ID was defined by the administrator of the batchprocessors.
Upon pressing the "Continue" button, OpenCA attempts to install the certificate into the user's browser. The screens presented to the user depend on the browser being used.
By pressing this link the user is presented with a screen displaying the session server and client certificate details. In most cases this screen will only display the details of the web server certificate used to secure the session (as this screen is not usually access via pages requiring client side authentication).
The user is offered the opportunity to "Sign" a set of data to test the client certificate. Upon pressing the "Sign" button the user is asked to choose the certificate they wish to use to sign the test data. Once they have chosen their certificate they may be asked to enter the pass phrase securing their private keys (this depends on how the user installed the certificate and private keys during key generation time). Once they that completed this the results of the signing process are displayed.
This screen gives the user the opportunity to revoke their own certificate. To do this they need to fill in the form and press "Continue". The Certificate Serial number can be obtained by examining the certificate (using browser functions) or by looking up the certificate in the valid certificates list. The CRIN code was sent to the user at certificate creation time.
This set of options provides the user with lists of certificates in various states, valid, expired, revoked and suspended. It also provides an interface for the user to search for a certificate.
Following this link the user is presented with a screen displaying all valid certificates. The screen shows 20 certificates at a time. The user can scroll through the valid certificates by using the "Extra References" link in the top right of the screen.
For each certificate the screen shows:
The serial number of the certificate
The common name associated with the certificate
The date and time the certificate was issued
The email address in the certificate (the user can click this link to mail the certificate holder)
The type of certificate (e.g. User)
A user can view the content of a certificate by clicking the serial number of the certificate they wish to view. OpenCA presents a screen displaying the certificate details. At the bottom of this screen are two new links where the user can download the certificate (and install it into their browser) or initiate the revocation procedure (in order to do this the user must have the CRIN number for the certificate being viewed, this number is presented to the certificate holder at certificate creation time, so only the certificate holder can revoke their own certificate).
Clicking this link shows the user a list of all the expired certificates. The screen shows 20 certificates at a time. The user can scroll through the expired certificates by using the "Extra References" link in the top right of the screen.
Clicking this link shows the user a list of all the revoked certificates. The screen shows 20 certificates at a time. The user can scroll through the revoked certificates by using the "Extra References" link in the top right of the screen.
Clicking this link shows the user a list of all the suspended certificates. The screen shows 20 certificates at a time. The user can scroll through the suspended certificates by using the "Extra References" link in the top right of the screen. Suspended certificates are certificates that have had the revocation process started but not yet revoked by the Certificate Authority.
This link provides a screen to enable users to search for a certificate on the system. The screen allows the user to search based on the criteria of name, email or distinguished name. Wild cards are allowed (e.g. Chris*) in each of the fields. You do not have to fill in each of the fields for the search function to find a match, but the more search data you enter the finer the granularity of the search.
This section displays outstanding requests. New certificate requests and revocation requests can be displayed.
Following this link the user is presented with a list of all the current certificate requests at the Registration Authority. The screen shows 20 requests at a time. The user can scroll through the list by using the "Extra References" link in the top right of the screen.
Following this link the user is presented with a list of all the current certificate revocation requests at the Registration Authority. The screen shows 20 revocation requests at a time. The user can scroll through the list by using the "Extra References" link in the top right of the screen.
This section describes the registration authority interface to the OpenCA PKI. From these screens an RA Administrator can manage certificate requests, view certificate information and manage the RA server.
The user is first asked to authenticate themselves to the RA, depending on the configuration, this authentication may be nothing, username and password or by certificate.
Each one of the headings below coresponds to tab accross the top of the default RA screens.
Pressing this link takes the user to the RA Node interface. From here the RA user can control data flow to and from the RA.
Pressing this link takes the user to the LDAP Administration interface. From here the RA user can control the import and deletion of data from the LDAP Directory (if it is configured).
This tab list functions that can be performed on Active Certificate Signing Requests, i.e. requests from users for a certificate.
This link shows new CSRs at a specific Registration Authority with a certain Level of Assurance (as specified by the user at certificate request time). The RA Operator chooses the RA and LoA.
The screen shows all the new CSRs.
Each one of the requests must be processed in turn. By clicking the serial number of the request the operator is presented with the details of the request. Four options are then available to the RA Operator:
Pressing this button allows the RA User to edit the details of the request. The editable fields are; Subject alternative name (this is usually defaulted to the supplied email address, but can contain other fields), Subject (or the DN) and Role (or certificate type).
Pressing this button allows the RA User to approve the request and use a certificate to sign this approval. Upon pressing the button the RA User is presented with a list of certificates with which to sign the request approval. Note, if the requests are going to be processed on the CA as a batch process, then each request must be signed with a valid RA certificate (signed by the certificate authority).
Pressing this button approves the request. Note, this can potentially be dangerous as the CA Administrator will have to make a trust decision to process the request or not. If the approved request was signed by a valid RA cert then this decision is unnecesary.
This screen displays any re-newed certificate siging requests. The list of options is the same as the "New" function.
A user can initiate their own certificate revocation or it can be initiated by an RA Operator. This screen shows Certificate Revocation Eequests in various states.
This section shows new certificate revocation requests. The RA Operator can process them by clicking on the CRR serial number.
Pressing this button allows the RA User to approve the revocation request and use a certificate to sign this approval. Upon pressing the button the RA User is presented with a list of certificates with which to sign the request approval. Note, if the requests are going to be processed on the CA as a batch process, then each request must be signed with a valid RA certificate (signed by the certificate authority).
Pressing this button approves the revocation request. Note, this can potentially be dangerous as the CA Administrator will have to make a trust decision to process the request or not. If the approved request was signed by a valid RA cert then this decision is unnecesary.
This section shows CRRs that have been approved and exported to the CA.
This tab allows the RA Operator a different view of CSRs, CRR, User certificates, CA certificates and CRLs.
This link displays the user submitted requests and enables the RA Administrator to presses them.
The following lists of certificate requests can be displayed.
New
Renewed
Pending
Signed (waiting for additional signature)
Approved
Archived
Deleted
This link displays the user submitted revocation requests and enables the RA Administrator to presses them.
The following lists of certificate revocation requests can be displayed.
New
Pending
Signed (waiting for additional signature)
Approved
Archived
Deleted
This link displays information about certificates in the PKI.
The following lists of certificates can be displayed.
Valid
Expired
Suspended
Revoked
This link displays information about CA certificates in the PKI.
The following lists of CA certificates can be displayed.
Valid
Expired
This section contains RA Operator utilities.
This allows the RA Operator to search for a specific certificate based on Name, Email, DN or Role.
This allows the RA Operator to search for a specific certificate signing request based on Name, Email, DN or Role.
The RA Node is the interface used to control RA operations that deal with external interfaces, for example exporting request data to the Certificate Authority.
This section lists the OpenCA components.
This link takes the user to the main CA Server page. This link is only available if the CA is accessable. In the normal OpenCA configuration the CA is offline and so this link will fail.
This section takes the user to a screen to manage the LDAP server, if one has been configured.
Pressing this link takes the user to the OpenCA public interface. This interface is described elsewhere in this document.
This section lists the options available to configure and maintain the RA Node.
If the PKI has been configured with Crypto Tokens holding the CA provate key in an online mode, then this link will stop the token deamon.
This screen is used to set up your OpenCA RA. It is intended that the screen be used once and once only. There are two links:
The RA Administrator should press this link to run the data base initialisation script. Note if you run this script on an existing data base then you are likely to loose all existing data. Be careful !
This link is used to exchange data with other areas of the PKI infrastructure (e.g. CA). Depending on your implementation of OpenCA, only some of the following sections will apply.
It is unlikely that there will be a lower level of the hierachy at the RA.
It is unlikely that there will be a lower level of the hierachy at the RA.
This section is used to download data from the CA to the RA. In order to use this section, data must have already been exported from the CA. This data is usually stored on a floppy disk. Upon clicking any of the following links the user is prompted "You need to provide a support to proceed (depends on your configuration). Are you sure you want to continue ?". This means that you need to have read access to the device that the exported data is on (e.g. the floppy drive).
Pressing this link imports all the data that has been exported from the CA into the RA.
Pressing this link imports only the certificate data from the CA into the RA.
Pressing this link imports only the configuration data from the CA to the RA. This is data like user roles (certificate types).
This section enables the RA Administrator to export data to the export device ready for import to the CA. Upon clicking any of the following links the user is prompted "You need to provide a support to proceed (depends on your configuration). Are you sure you want to continue ?". This means that you need to have write access to the device that the exported data going to be written to (e.g. the floppy drive).
Note, the export of data is in the form of a delta, i.e. only new or modified data is exported. It is an administration task to modify this behaviour.
This section allows the RA Administrator to backup and recover the OpenCA RA database. It is good practice to perform a data base backup regularly.
Pressing this link backs up the database to the export device. Upon clicking the link, the user is prompted "You need to provide a support to proceed (depends on your configuration). Are you sure you want to continue ?". This means that you need to have write access to the device that the exported data going to be written to (e.g. the floppy drive).
Pressing this link configures the data base for import of data. If you are rebuilding the RA then it is important to press this link.
General utilities for the RA Operator
Pressing this link sends the "New User" emails out to new users. These emails tell the users that their certificates are aready for collection and gives them a link to the public interface to collect their certificates.
Pressing this link sends new users an encrypted CRIN mail. The CRIN mail contains a pin code that the user must enter when revoking their own certificates. The user should be able to decrypt the message as they would have created the private key during their certificate request process. The message is encrypted using the public key in the certificate request.
This set of screens controls the uploading of data to the LDAP Directory (if there is one configured).
Pressing this link uploads the CA certificate to the LDAP server. The main screen shows this process and indicates the upload status and success or failure.
Pressing this link uploads the current valid certificates to the LDAP direcrectory and removes revoked certificates from the directory. The main screen shows the status and indicates success or failure.
This link displays a main screen showing the CA certificate. Clicking on the serial number of the certificate displays a new page showing the certificate details.
As this page is from the LDAP server, there are 4 LDAP related options:
Allows the user to enter a modified DN for the certificate and then publish it to the LDAP directory with the modified DN.
This link displays a list of the valid certificates. Clicking the serial number of a certificate displays the certificate details.
As this page is from the LDAP server there a 4 LDAP related options:
Pressing this button allows the user to enter a modified DN and then upload the certificate to the LDAP directory with the modified DN.
This link displays a list of the expired certificates. Clicking the serial number of a certificate displays the certificate details.
As this page is from the LDAP server there a 4 LDAP related options:
Pressing this button allows the user to enter a modified DN and then upload the certificate to the LDAP directory with the modified DN.
This link displays a list of the suspended certificates. Clicking the serial number of a certificate displays the certificate details.
As this page is from the LDAP server there a 4 LDAP related options:
Pressing this button allows the user to enter a modified DN and then upload the certificate to the LDAP directory with the modified DN.
This link displays a list of the revoked certificates. Clicking the serial number of a certificate displays the certificate details.
As this page is from the LDAP server there a 4 LDAP related options:
Pressing this button allows the user to enter a modified DN and then upload the certificate to the LDAP directory with the modified DN.
This link displays a list of the certificate revocation lists (CRLs). Clicking the Ver (***is this a bug, I think you should click the CRL serial number ***) of the CRL displays the CRL details lists the revoked certificates.
Clicking on the serial number of a revoked certificate displays the certificate details screen in the same way as pressing the "View Certificates Valid" link.
As this page is from the LDAP server there are 2 LDAP related options:
Table of Contents
The initialization of the CA consists of three phases and can only be executed once. The first phase is mandatory. It initializes the CA itself. The second and third phase are optional. They create the first two certificates. Phase two creates the first certificate for an operator and phase three creates the certificate for the online web server. First we describe phase one and then we describe phase two and three in one section because they are nearly identical.
The first phase of the initialization of OpenCA is used to setup all the cryptographic mechanism which are necessary to run a CA. This include things like a private key, a certificate signing request (CSR), a CA certificate and the certificate chain of the CA. If you completed this phase successfully then your CA is ready for operational usage. The other phases only do some optional tasks.
This link initializes the database. If you use OpenCA::DB then the backend consists of DBM files. If you use OpenCA::DBI then the backend consists of an SQL database. Today we support MySQL, PostgreSQL, IBM DB2 and Oracle. If you need another database then please contact us.
Here we do the keygeneration for the CA. You have to enter the algorithm for the key itself, the algorithm for the encryption of the key and the length of the key. If you entered all the cryptographic paramters then you must enter the passphrase if you use an software token from OpenSSL. If you use a hardware token then you must active this token by the appropriate action.
If you start creating a new certificate request for the CA certificate then you will be prompted for several informations which will be needed to create a certificate of the style “emailAddress=...,cn=...,ou=...,o=...,c=...”. After you entered all the data OpenCA will display the complete subject of the request. This is the last time when you can modify the subject. If you need another style e.g. dc style then you can enter in this field a subject of your kind. After you entered all parameter for the request the private must be activated (usually via a passphrase).
There are two general options for a CA certificate in OpenCA. You can use the CA as a root CA then you have to create a selfsigned certificate or you can setup a sub CA then you have to go to another CA and let it sign your request. Both variation need a different handling.
If you want to create a new root CA then you have simply to
create a new selfsigned CA certificate. This is much more
simple then to setup a new sub CA but it is more dangerous.
Before you create a new CA certificate please check
OPENCADIR/etc/openssl/openssl.cnf
. The
extension are in the section v3_ca
. It is
highly recommend to set the option
subjectAltName
explicitly.
If you click on the link then you have only to activate the
private key and the rest will be done automatically.
We talk about a “root CA” in this section but the CA which issues the CA certificate for the new CA which is a sub CA has not to be a root CA. We only use this term to have to different unique namespaces for the CAs in this section.
First you have to export the request to the root CA which
has to issue your CA certificate. OpenCA will create a tar
file on your export media. This tar file contains a file
careq.pem
. This file is your request
in PKCS#10 format. The encoding is PEM (base64). Please
go with this request to the root CA and follow its
instruction for request processing.
If the root CA issued a certificate for the new sub CA then
you have to create a new tar file on your import media.
The tar file must contain the file
cacert.pem
which is the new CA
certificate. If you click on the link for th import of the
new CA certificate then OpenCA copies the file to all
necessary places.
The last steps can also be done on the interface for the
nodemanagement but it is a good idea to do it during the
intialization to get a consistent state. The rebuild of the CA
chain is necessary to verify digital signatures correctly.
If you want to setup a sub CA then you must add all CA
certificates of the CA chain in PEM format to the directory
OPENCADIR/var/crypto/chain/
before you
rebuild the chain.
The really last step is the export of the configuration to the online server(s). The most OpenCA users ignore this step and handle all the communication between the different nodes of the PKI hierarchy via the interface for the node management. If this is you first OpenCA usage then you should export the configuration and import it into the online server.
These two phase are used to create the first certificates of the infrastructure. You can create these certificates via a seperate RA and public interface too but usually the people want only to setup one online RA and this RA should run with https. So they need at minimum a certificate for the RA's web server. If they use an online RA and it should be a secure solution then the operators must sign a request if they approve a request. So they need an initial user certificate. If the operators have to login via a digital signature and not with login and passphrase then the first user certificate is mandatory. If you compare the two screenshots then you will see that only the names of the forms differ.
Both phases use the same process like for every normal certificate signing request. Sou can simply read the user guide section where the different requests and their handling will be described. The phases consists of four steps:
This link uses the automatic browser detection of OpenCA so the key and request will be generated by the browser. If you want to use smartcards for the user certificates then you can create the first keypair on a smartcard too if you use the PKCS#11 or CSP drivers of the vendor. It is recommended to use the role “CA operator” for the first user certificate.
Like editing on the RA. See Section 3, “CSR Handling - a request HOWTO” for more details.
Like the issue link if you view a certificate on the CA. See the request handling in the user guide.
See Section 4, “Certificate Handling” for an explanation of the options.
After you setup an offline CA you usually wants to setup one or more online servers to publish certificates and CRL, receive requests and offers a lot of other services to your customers. To do so you have to install the necessary components plus a so called node interface. This node interface is used to manage the communication in the PKI hierarchy between the different servers and to manage the server itself. If you initialize such an online component then you have to use this node interface.
After you logged in you have to go to and then to . First you should initialize the database and then you have to import the configuration from the CA or any other server which is at a higher position of the PKI's hierarchy. The last step also creates things like a working certificate chain.
Usually an administrator or an user suddenly needs more security or is shocked how high does it's risk be if he wants to request a certificate. Then the customers try panically to request a certifcate and don't know how. This is the moment when you should be ready to present them a small explanation how the certification works. We want to give you here an overview how the complete process of OpenCA works. You can simply use this as a base for your user information. The explanations are divided into five plus two parts:
how to submit the new request
how to edit a request
how to approve a request
how to issue a certificate
how to enroll a certificate
how to delete a request
explanations of the input fields
Please remember that a request is more correctly a certificate signing request. There is a second request type too - the certificate revocation request (CSR).
X.509 certificates are very well known and the format is really established but in the first years there are some problems with the requests because there was no simple protocol to submit requests via clientinterfaces. The result is that there is one standardized format for requests and one proprietary format too. The standard request format is PKCS#10 which is supported by all servers (e.g. web, mail, VPN) and Microsoft Internet Explorer. Netscape developed an own format called SPKAC. This proprietary format is used today by Mozilla and Opera. Surprisingly there is software available which uses certificates and private keys to encrypt data like emails or https-connections but is not able to create a private key or a request. Such a software is KDE's konqueror today. This software simple want to load prepared private key and certificate.
If you think “Wow what a complex area!” then please are not shocked but I don't talked about the request generation until now. The request is only the data but we need something like a procedure for creation and transport of the request too. The browsers know two ways today - Microsofts CAPI (Microsoft Internet Explorer) and Netscape's tag “keygen” (Mozilla, Netscape, Opera). A third way is the way of konqueror which expects we do all for it. A second big area are servers. Some servers like Apache has the same behaviour like konqueror. So you have generate the keys and requests by hand or go to the PKI's public interface and let the PKI does the job. Some other servers like VPN server can generate private keys and requests by itself. Servers and VPN clients support in contrast to webbrowsers two transport protocols - the manual way and SCEP. Usually you can copy the request to a floppy and go to your PKI or you use SCEP to handle all the communication with the PKI automatically.
Now is the moment to calm down and realize that we try to cover all these aspects and you have only to know what you have for a software and what you want to do. The following sections describe the steps for the different kinds of requests and submission technologies. If you have an Internet Explorer or need a certificate for Microsoft Outlook (express) then you must only read the section for the Microsoft Internet Explorer and so on. If you don't know what to do then simply use the interface with the automatic browserdetection.
You have to enter all the informations on the first screen. Please use only keylengths of 512, 1024 and 2048 bits. Other keylengths are actually not supported. It is highly recommended to use at minimum 1024. We recommend that you use 2048 bit because of the last theoretical papers on RSA algorithm. 1024 bits are still safe but it can happen that 1024 will cracked in the next years. If you entered all informations correctly and click ok then you will see a second screen. If you do something wrong then you will see the first screen again.
The second screen shows you the entered informations again. Please check the details. If you click ok then your Internet Explorer will start to generate a new key and request. If the operation succeeds then OpenCA will show you a success page with the serial number of the request. Please remeber this serial before you get in touch with the registration authority.
SPKAC request are used by Netscape, Mozilla and Opera. All three browser support the special tag keygen which is used to generate the private key and the request.
On the first screen you have to enter all the informations. The keylength can be ignored because you must select it on the second screen again because there is no way to set a default. If you entered all informations correctly and click ok then you will see a second screen. If you do something wrong then you will see the first screen again.
The second screen shows you the entered informations again. Please check the details. Now you have to select a keylength. It is highly recommended to use at minimum 1024. We recommend that you use 2048 bit because of the last theoretical papers on RSA algorithm. 1024 bits are still safe but it can happen that 1024 will cracked in the next years. If you click ok then your browser will start to generate a new key and request. If the operation succeeds then OpenCA will show you a success page with the serial number of the request. Please remeber this serial before you get in touch with the registration authority.
There are two methods to submit a pregenerated PKCS#10 - the webinterface or SCEP. SCEP will be covered completely seperate from this section. If you have a PKCS#10 request then you can upload it to the PKI's database. If it is rejected with an errormessage related to the subject (the distinguished name) of the request and you don't know what you are doing wrong then please contact your registration authority. It is possible configure some restrictions to enforce usable requests.
This is not a real request type. It is more a webformular that send an information to the RA that you want a smartcard. The request does not contain any cryptographic informations. It is simply used to collect some first informations about you.
This technology uses the same webpages like Microsoft Internet Explorer or Mozilla but there is a big difference. The browsers will generate the private key and request by themselves and only send them to the server but this interface will only send your entered data to the server and the server will create the key and request. You can use this technology for exaple if you need a certificate for a server and you have absolutely no idea from PKIs or you have a browser like konqueror which can use certificates and keys but is not able to generate requests.
On the first screen you have to enter all the informations. If you choose a keylength then it is highly recommended to use at minimum 1024. We recommend that you use 2048 bit because of the last theoretical papers on RSA algorithm. 1024 bits are still safe but it can happen that 1024 will cracked in the next years. If you entered all informations correctly and click ok then you will see a second screen. If you do something wrong then you will see the first screen again.
The second screen shows you the entered informations again. Please check the details. If you click ok then your data will be send to the server again and the server creates a private key and a request for you. Please remember the PIN. It is used to protect the private key. The databases on the server have no backup from this PIN! If all operations succeeds then OpenCA will show you a success page with the serial number of the request. Please remeber this serial before you get in touch with the registration authority.
The automatic browserdetection is used to determine which mechanism should be used to handle your kind browser. Please use this only if you need a personal user certificate because it really difficult to extract a certificate for a server from a browser and convert it to an appropriate format. If you use an Internet Explorer then please read section for Microsoft client requests. If you use Mozilla, Netscape, Opera or a Mozilla derivate like Phoenix or Firebird then please read the section for SPKAC requests. All other user browser will covered by the section for serverside key and request generation.
There are several fields which you have to fill in. Here is an overview how they will used:
The input fields in this area are used to generate the subject of the certificate. The subject of a certificate is the name of the certificate.
There are a number of restrictions to the content of the subject fields, do not use characters "#", "," and ";".
These informations are only used internally. The entered data is stored in the database of the PKI but they are not part of the certificate. These fields are usually used for things like telephonenumber, addresses and personal IDs.
This is your role in the PKI. A role has several effects. First it decides which extensions will be added to your certificate. The extensions define for applications the certificate can be used. Second it can be used to manage the access rights in the PKI.
Deprecated. Will be replaced by LOA.
The PIN will be always hashed and stored in the database. It is planned to use the PIN to identify the issuer of the request because only he can know the PIN. This is not implemented today but the PKI can be configured to use this PIN to allow you start a revocation if you lost the private key and/or certificate. The PIN is called CRIN in this case.
If you use serverside key and request generation then this is the PIN for the private key. Only you know this PIN!
This is the used length of the private key. Please check this value on the second form too because some browser like Mozilla lost this information if OpenCA asks for your approval for the key and request generation.
If a user submits a request then this request must be verified by an operator. The operator can look at the request by using the request lists or search the database. If he wants to edit the request then he has to click on edit.
It is possible to change four things in the edit form.
The subject alternative name contains additional informations for other softwares which use the certificate but have not the private key. Therefore this alternative name contain informations like emailaddresses, DNS names, IP addresses and more. Emailaddresses are important if you use this certificate for emailencryption because the most clients support only S/MIME v2 and this requires an emailaddress in the certificate. DNS names and IP addresses are useful if you use the certificate for SSL servers or VPNs.
If you use a patched OpenSSL 0.9.7 or an OpenSSL 0.9.8+
to issue certificates for Microsoft Smartcardlogin then
you have to edit the configuration of the RA. You must
add the option for othername to
CSR_SUPPORTED_SUBJECT_ALT_NAMES
and you
have to set the correct number of available options at
CSR_DEFAULT_SUBJECT_ALT_NAME_FIELDS
.
This is the name of the certificate. The least significant information is at top. It is the same order like in RFC 2253 (LDAP string representation).
This is the requested role. The RA operator must set the correct one.
These are informations which will assembled only for internal usage. Here you can store such things like contact addresses and special phonenumbers.
If a RA operator is sure that all informations are now correct then he can approve the request. You can do this with and without a digital signature. The digital signature secures the request against manipulation after the approval.
If you are on the CA interface and you look at the request then you can only issue a certificate from the request or delete the request. If you click on issue then you have to enter the passphrases or whatever your OpenCA needs to unlock the private key. That's all.
Please read the section which describes the certificate handling to understand how certificates will be enrolled to the user. If you use a SCEP device then please read the SCEP section.
If you look at a certificate signing request on the RA or CA then you have the option to delete the request. This is necessary if a request has to be rejected by a PKI. If you click on delete then you will be prompted for an additional ok. If you want to make a deleted request valid again then go to the deleted request and renew it.
This section describes all the things which you can do with a displayed certificate.
You can find a certificate with two methods. The first method is search. Go to
--> . You can enter some parameters in the displayed search form. The form only accepts wild cards if you use a SQL database. If the search succeeds then you can choose the certificate which will be displayed.The second method is a little bit more “stupid”. Go to -->
and try to find the appropriate certificate in the lists. You can navigate by using the links in lineExtra References
.
You can directly download a certificate into your browser by entering an appropriate serial number. You must know the serial number of the certificate, of the request or you ID in the batch processors. The browser will be automatically detected by the software. Please remember that this method only works if you generated the private key with the browser and the private key is still in your keystore on computer.
There are three different ways to download a certificate. You can download passive data, or you can download the private key and the certificate or you can install the certificate of another user. If you already have the private key and you want to install a new certificate in your browser then please use the direct download, because this is the only software part which sends special HTML-pages for direct certificate installation.
If you only need a certificate in a special format then can choose the format and click on
. The certificate will be send with an appropriate MIME type which prevents browsers from installation. You can save the certificate on a disk and you can do what you (or the policy) want to do with it.If you want to download a certificate and the private key
there are two possibilities. If the operation is allowed
on your interface and the configuration switch
REQUIRE_PASSWD_PUBLIC
is set to
NO
then you can click on download.
If you need the key in a format different from PKCS#8
then you must enter the passphrase to convert the private
key. After this you will receive the key and certificate
and you can save them.
If the operation is allowed
on your interface and the configuration switch
REQUIRE_PASSWD_PUBLIC
is
set to YES
then you must go
to your RA Operator and ask them to set a
passphrase. We do this to avoid denial of service
attacks against the private key of a user. It is
strongly recommended do delete the passphrase
after a short period of time and to generate the
passphrases with things like openssl
rand. User or admin
“generated” passphrase are often not
really secure. If the admin for this certificate via
the RA interface then you can go again to your
interface and download the certificate and private
key. You have to enter the passphrase for the
private key first and then the software will ask you
for a second passphrase to grant you access to the
export command. If you downloaded the key then
please inform the RA Operator and ask him to
remove the passphrase to avoid denial of service
attacks against you private key.
Sometimes you need a certificate of another user
who has never sent you a signed mail. If you have
a normal installation with LDAP support then you
can search the certificate in the directory. There
are installations where this service is not
available. In this case you can go to the
certificate page and if the appropriate
functionality (INSTALL_CERT
)
is activated in the configuration then you can
click on install, and the certificate will be
automatically installed in your certificate
store. After this you can use it to encrypt
emails.
The OpenCA team introduced with version 0.9.2 support for SCEP. SCEP was developed by Cisco. The protocol is usually used by VPN products like routers, switches and software clients to submit requests, download certificates and CRLs. We will document here the configuration for known systems to work with OpenCA. The configuration of OpenCA's SCEP service is described in the administration guide Section 8, “SCEP”.
SSCEP means Simple SCEP client for Unix. It can be used to request certificates if a device doesn't support SCEP but you want to use SCEP. Usually SSCEP is used to test SCEP daemons to work properly.
Our example uses the following configuration
(sscep.conf
):
Example 6.1. SSCEP configuration
# URL of the SCEP server. URL http://scep.pki.openca.org/cgi-bin/scep/scep # This is one is needed with all operations. CACertFile ./ca.crt-0 # Possible values: yes or no. Verbose yes Debug yes # Display fingerprint algorithm (md5/sha1) FingerPrint md5 # Private key created with mkrequest PrivateKeyFile ./local.key # Where to write successfully enrolled certificate LocalCertFile ./local.crt # Certificate request file created with mkrequest CertReqFile ./local.csr # Poll periodically for pending certificate (seconds) PollInterval 60 # Maximum polling time MaxPollTime 28800 # Maximum polling count MaxPollCount 256 # Certificate serial number (decimal) GetCertSerial 1 # Write certificate as GetCertFile ./cert.crt # Write CRL as GetCrlFile ./crl.crl
First you have to download the CA and SCEP certificate from the SCEP
server. You can do this with the command
sscep getca -f sscep.conf.
Please check that the CACertFile is correct. Sometimes the
SCEP certificate is ca.crt-0-0. Now you have all what you need to
request a certificate. local.crt
must contain
your request and local.key
should contain your
private key. To avoid that your CA admin think that your are an
attacker please poll not faster then all ten minutes (600 seconds).
You can run now sscep enroll -f sscep.conf. The result is some debugging output and sscep starts polling until it reaches it's maximum poll time, receives an errormessage from the SCEP server or downloads the certificate successfully. That's all. No magic but a really simple and efficient client interface.
If you don't want polling e.g. if you use SCEP for web server requests then you can start the enrollment and kill the process when SSCEP starts polling. If you know that the certificate is available then you simply start the enrollment again. SSCEP tries to send the request again but OpenCA detects that the request is already present in the database. So OpenCA reports a successful submission of the request and the first polling of SSCEP ends with the submission of the new certificate. If you think now that this is a good program to implement a batchprocess to generate smartcards etc. then you are right (but OpenCA has batchprocessors too) :)
OpenCA's SCEP service is tested with NetScreen NS-208. NetScreen's SCEP implementation sends SCEP messages in base64 but without any newlines. Now OpenCA can handle this too.
First you have to install the complete CA chain. You have to go to
objects
and then to certificates
.
Here you must set the option Show
to
CA
. Now you can upload the CA certificates
via browse
and load
.
After you
uploaded the complete chain please go to the end user CA and click
on Server Settings
. The interface is a little
bit mistakable because it display the issuer and in this field you
find the link to configure the CA. RA CGI
and
CA CGI
must be set to OpenCA's SCEP interface. The
address is something like
http://scep.mypki.org/cgi-bin/scep/scep
.
If you want to be consequent then please check the
advanced settings
to be correct for your
environment. It is recommended to set at minimum the field
Certificate Renew
to seven days. Finally click on
OK
to save the settings. This can take some time.
Now it's time to make the request. Change Show
from CA
to Local
and
click on New
. Enter all required informations
and choose at minimum a keylength of 1024 bit - smaller keylengths
are a security risk. If you finished then click
Generate
to create the key and request. Due to the
slow hardware it can take some time. If you see the request then
select the checkboxes Automatically enroll to
and Existing CA server settings
. Select the
appropriate CA which you configured for SCEP and click
on OK
. This will submit the request.
If the certificate was issued go to the web interface of your
NetScreen box. Go to objects
and then to
certificates
. Here you must set the option
Show
to Local
. Click on
Retrieve
to check for the certificate.
The SCEP compliance with OpenCA was tested with F-Secure Management Agent 5.02 (FSMA) and F-Secure VPN+ 5.43 (Gateway and Client).
If you want to use SCEP with VPN+ then you must set the appropriate policy in FSMA. Go to /Settings/Certificate Handling/Enrollment/Active protocol. There you can choose which enrollment protocol you need. Simply click on SCEP to activate SCEP.
After the activation of SCEP you have to configure it. The configuration settings can be found under /Settings/Certificate Handling/Enrollment/Protocol Settings/SCEP. The options are
The value should look like
http://scep.mypki.org/cgi-bin/scep/scep
.
This is the default SCEP gateway of OpenCA.
This parameter is not used by OpenCA. You can ignore it.
We don't use the challenge password today. So you can ignore it too.
The RA interface of OpenCA allows the editing of all
parameters of the request except the public key. So it
is not important to select the right option here.
Nevertheless it is recommended to use
Only in SubjectAltName extension
because this avoid the duplication of data in the
subject of the request (distinguished name). VPN+ adds
the VPN+ identity via the common name to the subject
and this usually duplicates the entries in the DN. The
recommended option let VPN+ create a CSR which has the
identity in the subject (DN) and the attributes like
emailaddress, IP address and DNS name in the subject
alternative name.
That's all of the SCEP specific stuff in VPN+. You can read this description with some more details in F-Secures documentation too. You can find some docs on your CD-ROMs and some old docs online at F-Secure.com.
The most OpenCA SCEP users are using PIX but nobody writes a good documentation until now :( We can only notice that it is sometimes necessary to create a new role which only includes the extension for the Netscape certificate type. It looks like the PIX is some kind of sensible for special extensions.
Please notice that the PIX has one major problem - it cannot scale for large installations. This does not mean that there is not enough performance but the box can only handle one CA. This mean that if your have to replace your old CA by a new one then the trust for all old certs breaks immediately. It is not possible to operate with two CAs at the same time. If you have a large company with thousands of employees then it is nearly impossible to replace all user certs at one time.
Timestamp: 2004-May-17
Table of Contents
The biggest problems of PKIs are like for all other big infrastructures the clients. The big advantage of PKIs is that all vendors accept the basic standards. The big disadvantage is that the area of PKI-enabled clients changes very fast and some vendors are in a very early implementation phase or have products with different behaviour.
If this sounds confusing then you can believe us it is confusing. Therefore you can find here some descriptions for the different browsers and the common problems with these browsers. If you have additional problems or other informations then please mail us. We try to support as many client as possible but sometimes a new release can outdate a complete description.
This section describes the behaviour of all spin offs from the original Netscape browser. This includes in the security area Opera because Opera includes the crypto stuff in the same way like Netscape. This doesn't mean that Opera includes code from Netscape or Mozilla. Opera only operates in the same way and therefore we included it here.
If you use a Mozilla-like browser and you want to request a
certificate from an OpenCA-based PKI then you have to go to the
public web interface and klick onto the User area. There you can
choose an option request a certificate
. After
this you can choose between several options
(see Figure 7.1, “Request a certificate”).
You can choose the link for the automatic browser detection or more error proof the link for SPKAC-based requests. SPKAC is a special format defined by Netscape. This format is used by Netscape, Mozilla and Opera.
Please fill in your data at the next page. If you submit your data and the software finds no mistakes then you will see the data again. Please check them carefully to avoid additional work for your registration authority. If you now submit the form again then your browser generates a new private key and creates a new request with this key. This request will be send by your browser to the web server from your PKI.
Please print the last page or at minimum write down the displayed information about the necessary procedure and the displayed serials. These serials are necessary to install the later issued certificate.
There were several problems with Mozilla and signing in the past. The most problems should be fixed if you are using Mozilla 1.7+ or Firefox 0.9.3+.
These versions of Mozilla support the same technology like the old Section 2.3, “Netscape 4”. You must have the complete CA chain of a certificate and you must trust the CA certificate. If these requirements are meet then you can use a certificate to sign an HTML form. This is completely done in Javascript. No additional plugins are required.
If you want to sign HTML forms with the old versions of the Mozilla browser then you must use Section 2.2.2.3, “SecClab”.
Secclab is a XPCOM Component that implements some PKI functions. This sounds a little bit abstract but it is the official description from http://secclab.mozdev.org. In fact SecCLAB is a plugin which implements form signing.
All Mozillas from 1.0 to 1.6 and the modern Netscape version
doesn't include the old crypto object from Netscape 4. This
statement is correct for all Firefox up to and including 0.9.2
because Firefox uses the core code from Mozilla. The
result is that your unable to sign form with these browsers.
OpenCA needs this functionality to protect approved requests
and to support X.509 based authentication. The plugin
implements a new object class which replaces the old function
signForm
.
If you want to download or install the plugin from the homepage of SecCLAB then please notice that you must take care about the correct version. There are especially for Linux two different version - one for gcc3 and one for the old binaries. All new distributions use gcc3. Debian woody uses 2.95 and 2.96 (ia64). Only hppa (HP PA-RISC) uses gcc 3.0 but the plugin was only compiled for i386. After the installation you you have to quit Mozilla and start again. Please really quit Mozilla and not only close the used window (usually Ctrl-Q do the job). If you don't find the appropriate plugin in the download area then please check the installation area. There was some confusion with the gcc3 version for Linux.
If you want to sign something with a certficate then you must install the complete CA chain and you must trust the CA. The important thing is that you must trust the CA for email signing. We do not know what the signing of some data has to do with email signing but without this it does not work.
Before the description starts please notice that the SMTP based replication of Microsoft ADS doesn't work with TTPs. Microsoft officially commits this problem but doesn't comment the reasons. If you don't use SMTP replication then you can TTPs without any problems except that Microsoft's automatic certificate renewal does not work of course.
The requirements are described in detail in the knowledge base article 291010. The most important things are listed here:
There must be at minimum one CDP.
The X.509v3 extension Key Usage must
include the options Digital Signature
and
Key Encipherment
.
The X.509v3 extension Extended Key Usage
must include
Client Authentication (1.3.6.1.5.5.7.3.2)
and
Server Authentication (1.3.6.1.5.5.7.3.1)
.
This is
necessary because domain controllers will be connected as
servers and establish connections as clients to other servers.
The X.509v3 extension Subject Alternative Name
must include the globally unique identifier (GUID) of the
domain controller and the DNS name of the domain controller.
The GUID is a 16 Byte long number which identify the object of
the domain controller in the directory (ADS). The OID of the
GUID in the othername is 1.3.6.1.4.1.311.25.1
.
There must be one extension with the BMP value
DomainController
inside. Microsoft has an
extra extension for this template name with the OID
1.3.6.1.4.1.311.20.2
. The extension is
referenced by Microsoft as
Certificate Template Name.
The X.509v3 extension Basic Constraints
should include End Entity
as
subject
(OpenSSL: CA=false) and no limitation
for the path length.
The subject of the certificate can be the directory path of the domain controller. This is only optional.
If you want to create such a certficate then you must create the key pair with RSA/Schannel CSP. The most simple way for doing this is to install Microsoft CA services and to issue a certificate request using a form, where you can choose this CSP and generate a key pair and certificate request in PKCS#10 format.
The settings for Key Usage, Extended Key Usage and the CDPs is
no problem with OpenSSL and we don't describe here in detail how
you can configure this in
OPENCADIR/etc/openssl/extfiles/domaincontroller.ext
.
The Basic Constraint and Subject stuff is the same like for
every other certificate too.
The Subject Alternative Name is the first difficult thing. OpenSSL 0.9.7 does not support othername. The solution is to encode the GUID in binary format into the subject alternative name.
You can find a detailed description of DER:
in openssl.txt
of the OpenSSL
documentation. 2.5.29.17
is the OID of the
subject alternative name. The example GUID is
01020304050607080910111213141516
. Please
remember that the GUID is allways a 16-byte string and the
numbers are hexadecimal encoded. A 16 is a decimal 22. The
DNS name is dns.somehost.de
(64:6E:73:2E:73:6F:6D:65:68:6F:73:74:2E:64:65
).
If you need more infos about ASN.1 encodings then please read
the ASN.1 specs. We only know that the 10 in front of the GUID
and the 0F in front of the DNS name are the length of the
values.
The next complicated stuff is the certficate template name but this is a fixed extension so we can copy it from already existing certificates :)
After you read all these complicated issues you understand perhaps why I can only recommend you to use OpenSSL 0.9.8 if you want to issue certificates for domain controllers. The biggest problem is the subject alternative name which can only be created in binary format which requires to change the extension configuration file of the used role for every issued certificate to encode the correct GUID and DNS name of the domain controller. Nevertheless you can do it with OpenSSL 0.9.7.
The first important note before you start with OpenSSL 0.9.8 - you cannot compile OpenCA with OpenSSL 0.9.8. You must install OpenCA with an OpenSSL 0.9.7 and then you must change the path of the OpenSSL binary from the 0.9.7 binary to the 0.9.8 binary. We know that this is not really comfortable but the header files changed from 0.9.7 to 0.9.8 in an incompatible way and we only migrate our sources if there is a 0.9.8 stable release.
The standard things like Key Usage, Extended Key Usage, Subject, CDPs and Basic Constraints do not change between the different OpenSSL versions. The certificate template name extension of Microsoft must be copied as binary too but it is like the other standardized extensions a static string.
The important improvement of OpenSSL 0.9.8 is the support for
othername
in the subject alternative name. The
new OpenSSL version allows a much easier specification of
proprietary extensions even if they are in binary format.
The GUID is the same GUID like in the example for OpenSSL 0.9.7.
The big question is now how can I put it into my OpenCA?
The answer is short. You have to choose otherName as attribute
for the subject alternative name at the RA interface and then
you have to enter
1.3.6.1.4.1.311.25.1;FORMAT:HEX,OCT:01020304050607080910111213141516
.
We know that it is a big problem for the most admins to enter
OIDs and cryptical things like OpenSSL formats. Therefore we
added some extra stuff for Microsoft only. You can use the
field MS_GUID
as a subject alternative name
attribute too. There you have only to enter the pure GUID in
HEX format. The rest is handled by OpenCA.
The requirements are described in detail in the knowledge base article 281245. Please note that we only list the requirements for the PKI system here. There are several other requirements to which you can find in the knowledge base article. The most important things are listed here:
There must be at minimum one CDP.
The X.509v3 extension Key Usage must
include the options Digital Signature
.
The X.509v3 extension Extended Key Usage
must include
Client Authentication (1.3.6.1.5.5.7.3.2)
and
Smart Card Logon (1.3.6.1.4.1.311.20.2.2)
.
The X.509v3 extension Subject Alternative Nme
must include the universal principal name (UPN) of the user.
A UPN is the like account@domain
(e.g.
john_doe@company.com
). It is important to
understand that this is no emailaddress. The OID of the UPN
is 1.3.6.1.4.1.311.20.2.3
. The format of the
UPN is UTF-8.
The X.509v3 extension Basic Constraints
should include End Entity
as
subject
(OpenSSL: CA=false) and no limitation
for the path length.
Please use the patch for OpenSSL which you can find at our ftp
server. The patch othername.tgz
includes a
fix for Microsoft's othername usage. The following documentation
only refers to this patched version of OpenSSL.
OpenCA's role User
already includes the
required extensions for Microsoft's smartcard logon.
Generally you can add the OID for secure email
(1.3.6.1.5.5.7.3.4) and other
things to such a certificate. The smartcard certificate is not
limited to logon. To be more Microsoft-like you can add a
certificate template name to the certificates like for domain
controllers. This is not required but it looks more nice :)
Now we have to deal with the UPN. This UPN must be placed in the
othername
of the subject alternative name. If
you use the patched option and the othername
is
supported by the RA configuration then you have to enter
1.3.6.1.4.1.311.20.2.3:UTF8String:account@domain
in the value field after you selected
otherName
. The keyword UTF8String
is case-sensitive. If you do all correct then you can now issue
a wonderful certificate for smartcard logon.
The only difference between OpenSSL 0.9.8 and 0.9.7 is the
handling of the othername. OpenSSL 0.9.8 support this by default
without patching. You have to enter
1.3.6.1.4.1.311.20.2.3;UTF8:account@domain
.
Please notice that OpenSSL 0.9.8 uses UTF8 and not UTF8String.
We know that it is a big problem for normal users to enter
OIDs and other cryptical things like which have nothing to do
with the UPN. Therefore we added some extra stuff for Microsoft
only. You can use the field MS_UPN
as a
subject alternative name attribute too. There you have only to
enter the pure UPN like an emailaddress. The conversion to
OpenSSL's format is handled by OpenCA.
The private keys of a user are stored in the profile of the user if no smartcards are used. This means that the private keys of any Microsoft client software like Internet Explorer or Outlook are automatically present after a successful login because they are part of the profile. This protects the keys too.
Table of Contents
This guide has only one goal, it should explain all core technologies of OpenCA. This is not a guide which explains how you setup OpenCA or what you have to do if your OpenCA installation doesn't work but this guide is a good source to understand the software design of OpenCA.
So if you think there is a core part of OpenCA and you don't understand why it is implemented in this way then it is a good idea to read this guide or to write a new section if you think it should be documented.
Table of Contents
OpenCA includes a lot of ideas and we think it is impossible to give a 100 percent complete overview about all technologies and ideas. The development model of OpenCA is an evolutional model. The community develops new features if they need them. There is no strict release schedule or big corporation in the background.
This guide also follows the evolutional development model. It contains descriptions of new technologies and core concepts of OpenCA. It is impossible for us to document every detail but we want to give all developers a chance to now what OpenCA is. The developers of stylesheets doesn't develop necessarly the code for the encryption tools but sometimes they want to how they can use encryption. In this case they can take a first look into the Tech Guide to understand the general concepts.
Today this guide includes the following topics:
Before we start with concrete technologies it is necessary to describe a general technology - slots. Perl's DBI is a good example howto integrate several database drivers without installing and using more modules then required. This is what we call slot technology.
If you have a module and it needs a module which provides a special interface then it calls a "supermodule" which loads the required module and return it or handle all requests for the user including the management of the loaded modules.
The general mechanism is really easy:
my $class = "OpenCA::Token::$name"; eval "require $token_class"; return $self->setError ($@, $@) if ($@); my $token = eval {$token_class->new (@_)}; return $self->setError ($@, $@) if ($@); return $self->setError ($token_class::errno, $token_class::errval) if (not $token);
I hope you see it is really simple to develop a slot concept :-)
The first step is to load the required class. This has the same effect like a statical "use" statement but "use" doesn't work with dynamic classnames.
The second step is the creation of a new instance. Perl's OO interface is really bad compared with python or others like ruby but it works and we have not the time to rewrite the complete code. After the second eval you have a normal object reference and a new "slot" is established.
Please remember these concept at every time you read from a slot based technology.
This is no real description of a technology. It is more a statement of the development team. All new configuration files will be in XML.
The idea of all the XML files is the possibility to edit the configuration with texteditors and XML editors. So it is possible to develop a graphical interface for the configuration of OpenCA. The real important statement is the DTD. Today there is DTD but every XML file must contain <openca> as root element. The element below the root element must describe the type of the configuration (e.g. <token_config>). This makes later integrations of different configfiles really easy.
Today we have no DTD but it MUST be written before version 1.0.
The idea for a cryptolayer was born when we started to introduce one new private key every two weeks during the design of the batchprocessor. OpenCA needs today up to four different private keys - CA, logging, batchprocessor, key backup. Additionally we need a cryptoengine for the generation of private keys in the script basic_csr.
We use here the above described Section 1, “Slotechnology” to manage all the different keys. There is a central class OpenCA::Crypto. This class can load and maganage so called "tokens". The software itself uses at every time a special token. Usually there is a default token $cryptoShell avaialable which is not associated to a special key.
If a script or function needs a special key then it calls getToken an receives a loaded object for the appropriate key. The key must be activated by useKey because some tokens can be used without an authentication if the private key is not used. Such a "key" is a software token which is handled by OpenCA::Token::OpenSSL.
The tokens can fallback to the defaulttoken if they don't implemenent a functionality. The token class OpenCA::Token::Empty is only a logical class which automatically falls back to the default token which must be specified in etc/token.xml.
Today there are two additional classes beside the classes for empty and software keys - LunaCA3 and OpenSC. The class LunaCA3 can manage SafeNet LunaCA3 and Luna SA (former Rainbow former Chrysalis-ITS) devices. This module supports the daemon and session mode too. These modes allow the activation of the module for a complete session or forever (daemon mode). The session ends if the user logs out. This is of course critical but the user decides what he want. The daemon mode activates the HSM and runs it until there is an exlicit shutdown for the HSM via an interface. The class OpenSC can use any OpenSC smartcard as a HSM but it can also be used to integrate every other PKCS#11 device if you don't use the keygeneration of this class. The keygeneration is specific for OpenSC. The other stuff bases on OpenSC's PKCS#11 engine for OpenSSL which uses the PKCS#11 module of OpenSC. You can replace OpenSC's PKCS#11 library by any other PKCS#11 library.
During the implementation of some new PKIs several users have serious problems because OpenCA has no own powerful accesscontrol and the implemented accesscontrol uses a proprietary modified Base 64 encoding for filenames. The result of this difficult usage was that the developer itself (me ;-) ) don't use this RBAC implementation. So we take the ideas from the first try and start a second XML based try :-)
The accesscontrol has one central configurationfile for every web interface in etc/access_control/. This configurationfile contains references to the token configration and to the access control list if such a list will be used. The file itself has the usual OpenCA structure:
<openca> <access_control> The complete configuration should be here. </access_control> </openca>
The problem of the old system was the high complexity therefore it was one main goal to made a modular design that is maintainable by different developers. The system consists therefore of four parts:
channel verification
login
session management
ACLs
Every step is a completely isolated pass except of the second and the third step which are unified in the second pass. It is necessary to develop a Section 1, “Slotechnology” which handles the sessions and the data which is stored in the sessions because the actual accesscontrol is only usable with web interfaces but in the far future there should be at minimum a Tk interface.
The different passes works like follows.
The exact configuration of the different passes is explained in the administration guide.
Logging is a feature which is required for all real security systems. So we see the missing logging as a real problem for OpenCA and so we added it during the 0.9 series. The logging itself based on a slotmechanism to support as many different logging technoligies as possible.
If you want to create a logentry then you have to create a log message first. Such a message is really simple to create:
Every message wil be build from one hashreference. This reference include values and references to other hashes. So you can set a value via this construct
$hash->{user_data}->{emailaddress} = "testuser@abc.com";
The hashreference must include a key CLASS which you can freely choose. This class can beused to filter messages and to decide which slot should be used.
The hashreference must include a key LEVEL which has to be from one of the following values:
EMERG
ALERT
CRIT
ERR
WARNING
NOTICE
INFO
DEBUG
These are the standard Unix levels for messages.
If you create the message then you have to call new with the parameter HASHREF:
OpenCA::Log::Message->new (HASHREF => $hash)
If you created a log message then you have to put them into an initialized log module. The module will be intialized by the main scripts like ca and RAServer. The rest will be handled by the OpenCA::Log module.
my $log_token = $crypto_layer->getToken ('LOG'); $log_token = $cryptoShell if ( not $log_token ); $log = OpenCA::Log->new (CONFIG => getRequired ('LogConfiguration'), CRYPTO => $log_token); $log->addMessage (OpenCA::Log::Message->new (HASHREF => $hash));
The first line try to load a eventually configured special crypto token for logging. The second line falls back to the default token. This behaviour is used to support HSMs which run in daemon mode and can be used to sign the log. The third line created the logobject and last line add the message.
The configuration of the logging will usually done in etc/log.xml. Today there are two different logger implementation Syslog and XML. You can specify accepted class and level for every configured logger. Class and level accept wildcards.
The Syslog logger needs the type of the used Perl module. You can choose between Sys, Net, and Unix. Today only Sys is really tested. Additionaly you can specify a prefix for every log message. This is useful if you filter your logs with logsurfer or other tools. Please specify a facility too. This allows you to specify the behaviour of the syslogd via syslog.conf. If you use Sys then please specify the used socket type too. I used unix and it works all other types don't work on my linux box but it is better to allow the full flexibility and let it to the user to decide what he needs and uses.
The XML logger only has one option - dir. It specifies which directory the logger should use to establish the needed structure.
Table of Contents
OpenCA was designed for very fast and radical interface customization. Costumization is understanded as changing the functionality of the interface not the design. The design should be changed via CSS. It was a design goal to allow the user to create an interface for his own special requirements. OpenCA comes with predefined interfaces like ca, ra, public, ldap and node management but the user can merge split or completey mix the functions. This chapter will describe the ideas behind an interface.
There are three mainparts of the customization - statical webpages which show the user all the available functions, the functions itself and links which are only available for some commands.
OpenCA includes also some support for hiding functionalities. The command viewCSR, viewCert and viewCRR can be configured to only show some seletected links. This is not a security issue because it only provides a functionality like linkhiding but it shows the user a consistent interface.
The first section will describe the design to give you chance to understand what is and why it is going on before we describe the possibilities of customization in detail.
Former versions of OpenCA like 0.9.0 and 0.9.1 use several sheets to support the different commands with templates for the displayed data. This was a nice effort for the first releases but it creates a lot of trouble in some important areas. If you have hundreds of templates then you must translate hundreds of templates. It is really difficult to extract all text informations from templates and you extract really many HTML tags what means that you have style informations in your language database. The next problem is the customization. If you have hundreds of pages and you want to change smething for better CSS stylesheets then you have to change hundreds of templates - really stupid work. So the CCC camp 2003 was the right place to overcome with this solution.
The new solution is the first output module of OpenCA - OpenCA::UI::HTML. It provides OpenCA with four output methods
libSendReply
libSendMenu
libSendStatic
several logging functions
Let us start with the easiest thing - point four. Nothing changed in this interface except of some classes for the stylesheet. All output pages of OpenCA support stylsheets now. This results in some problems with old Netscape browsers. All actual browsers like Mozilla, Microsoft Internet Explorer, Konqueror etc. have no problems. The old Netscape ones cannot interpret stylesheets correctly. They only output the pure links and not the nice tabbing style like the other browsers but it was time to make a cut and to take up the cudgels for the future (who will use Netscape 4.7 in one or two years?).
libSendMenu
creates and manages menus from
menu.xml
. You can configure the complete menu in
this single file. If you want to hide some complexity then you can
modify this XML-file. If you installed OpenCA from scratch then you
can search for the pub interface in this file and you will see the raw
structure. The DTD is like this:
It is not a real DTD. It is more a grammar. OpenCA support now menus with unlimited depth but you have to take care about the frames of course.
libSendStatic
is used by
getStaticPage. Here you can define the functions
which the user can see if you show him some statical information. The
interface supports NAME, EXPLANATION, TIMESTAMP and ITEM_LIST.
libSendReply
is more dynamic than
libSendStatic
. Please check the source code until
I checked in the docs tomorrow. It supports command panels and lists,
item lists, info lists and siginfo. If you need to build statical
pages then please use getStaticPage (see listCSRtype in this file for
an example).
OpenCA still includes some statical webpages which will be
created the command getStaticPage
. If
you want to change the content of a static page then please
change this file.
There are two possibilites for additional functions - you can write a new one or you can use and already existing one.
If you write a new command then there are two important things. First Every command include a function cmdsFilename. This is the function which is called by the main script if the function filename is requested. Second if you implement a script which has to handle different states then please write a function checkFilename which includes all the status checks to avoid that somebody can start statusinjections.
If you want to establish the new command in the sourcecode
then you must add the filename in
src/common/lib/cmds/Makefile
which installs the file. Additionally you have to
create a file
src/common/etc/rbac/cmds/filename.xml
.
This file has to contain the configuration for the
access control. Please add after this step default
entry to the access control list in
src/common/etc/rbac/acl.xml
If you want to add the new command directly in an
existing installation then install the file in
lib/cmds
and create the access
control file in etc/rbac/cmds/
.
Please don't forget to set a correct entry in the ACL
at etc/rbac/acl.xml
.
There are some scripts which are able to hide references. viewCSR, viewCRR and viewCert are such scripts. There is an option CmdRefs_viewCSR in the config file. There you can specify which functions should be available via this interface. Please remember that this has nothing to do with security! The access control will be handed the normal access control and by the installation of the scripts. Link hiding is no security. The following options are actually supported:
CmdRefs_viewCSR
EDIT
,
APPROVE
,
APPROVE_WITHOUT_SIGNING
,
ISSUE_CERT
,
ISSUE_CERT_NEW
,
ISSUE_CERT_RENEW
,
ISSUE_CERT_PENDING
,
ISSUE_CERT_SIGNED
,
ISSUE_CERT_APPROVED
,
DELETE
,
DELETE_NEW
,
DELETE_RENEW
,
DELETE_PENDING
,
DELETE_SIGNED
,
DELETE_APPROVED
,
RENEW
,
RENEW_ARCHIVED
,
RENEW_DELETED
,
GENERATE_KEY
CmdRefs_viewCRR
EDIT
,
APPROVE
,
APPROVE_WITHOUT_SIGNING
,
REVOKE_CERT
,
REVOKE_CERT_NEW
,
REVOKE_CERT_PENDING
,
REVOKE_CERT_SIGNED
,
REVOKE_CERT_APPROVED
,
DELETE
,
DELETE_NEW
,
DELETE_PENDING
,
DELETE_SIGNED
,
DELETE_APPROVED
CmdRefs_viewCert
INSTALL_CERT
,
LDAP
,
REVOCATION
,
SENDCERT
,
SEND_CERT_KEY
,
VIEW_CSR
,
TOKENHANDLING
,
MAIL
,
SET_PUBLIC_PASSWD
,
DELETE_PUBLIC_PASSWD
Perhaps a small hint - if you setup more than one public (or any other) interface then please check that configure_etc.sh only configures this interface. If you don't check this then configure_etc.sh will reconfigure the other public interfacers too and this will crash their paths.
The OpenCA project has the goal to develop a trustcenter software which can be used out of the box. We don't ship prepared policies but we want to support the user with simple installation and update routines. This requires that the user can install a RPM or DEB package and then configure the software.
Such a complex software like a trustcenter cannot be configured with one single configuration file but it is impossible to ask the user for the configuration of over 200 files. Therefore we develop a post-install method for a postconfiguration after "configure;make;make install". The result is a single file etc/config.xml and a script etc/configure_etc.sh.
The file contains all basic configuration options which must be changed by the user. You simply configure config.xml and then you run the script which configure all template files.
If a criticism asks now why it is not possible to have one configurationfile if config.xml manages the complete configuration then please remeber the word "basic". Several CA admins have to change the configuration of the extensions for example to support special softwarerequirements. So config.xml is a technology to create useful binary packages but it is necessary to have all the other configuration files to use the full flexibility of the standards.
Table of Contents
Table of Contents
Be warned - this is a developer documentation which only documents the possibilities and technical background of OpenCA ldap caode but this is not a howto or a user documentation.
Of course we will not list top here.
STRUCTURAL
country
device
inetOrgPerson (inherits from organizationalPerson)
locality
person
organization
organizationalPerson (inherits from person)
organizationalRole
organizationalUnit
AUXILIARY
dcObject
pkiCA
pkiUser
opencaUniquelyIdentifiedUser
opencaEmailAddress
opencaSCEPDevice
If you are missing a special attribute for your installation then please contact us (openca-users@lists.sf.net).
dc
c
o
st
l
ou
unstructuredName
unstructuredAddress
cn
sn
emailAddress
serialNumber
The following table describes what we use for objectclasses and attributes if we insert a node in the LDAP tree. The least significant component of the distinguished name is the "LSC of the DN". This doesn't document the used auxiliary classes.
Table 15.1. Schema usage
LSC of the DN | filled attributes | filled attributes if present | objectclass stack |
---|---|---|---|
dc | dc | top, dcObject | |
c | c | top, country | |
st | st | top, locality | |
l | l | top, locality | |
o | o | top, organization | |
ou | ou | top, organizationalUnit | |
unstructuredName | cn | unstructuredName, unstructuredAddress, serialNumber, st, l, ou | top, device, opencaSCEPDevice |
unstructuredAddress | cn | unstructuredName, unstructuredAddress, serialNumber, st, l, ou | top, device, opencaSCEPDevice |
cn | cn | cn, st, l, ou, mail | top, organizationalRole (opencaEmailAddress) |
sn | cn | cn, st, l, ou, mail | top, organizationalRole (opencaEmailAddress) |
emailAddress | cn | cn, st, l, ou, mail | top, organizationalRole (opencaEmailAddress) |
serialNumber | cn | serialNumber, o, ou, l | top, device |
If we add a node to the directory tree then we add at every time to the objectclass stack the classes pkiCA and pkiUser. This is perhaps not the cleanest solution but it is safe for every possible configuration. If we add a node with the class organizationalRole then we add the auxiliary class opencaEmailAddress if an emailaddress is present.
If we add a node for the subject of a normal certificate to the directory tree then we use modified objectclass stacks to support ldap browsers which search for a special emailaddress. We don't do this generally to return only a node of the tree if this node contain a certificate. This was a problem in several old releases of OpenCA.
Table 15.2. Schema usage for user certificates
LSC of the DN | filled attributes | filled attributes if present | objectclass stack |
---|---|---|---|
cn | cn,sn | mail, o, st, l, ou | top, person, organizationalPerson, inetOrgPerson |
sn | cn,sn | mail, o, st, l, ou | top, person, organizationalPerson, inetOrgPerson |
emailAddress | cn,sn | mail, o, st, l, ou | top, person, organizationalPerson, inetOrgPerson |
serialNumber | cn,sn | mail, o, st, l, ou | top, person, organizationalPerson, inetOrgPerson, opencaUniquelyIdentifiedUser |
If the distinguished name doesn't contain an emailaddress but OpenCA detects an emailaddress in the subject alternative name then we use this emailaddress.
addCertsLDAP (puts all valid certs to LDAP)
addCrlLDAP (puts all CRLs to LDAP)
importAllFromCA (via export-import.lib)
importCRL (via export-import.lib)
importCerts (via export-import.lib)
importCertsLDAP (puts all certs from the last import to LDAP)
importConfig (puts CA-certs to LDAP)
updateCACertsLDAP (update the CA-certificates on the ldap server)
updateCRLonLDAP (writes the most actual CRL to LDAP)
updateCertsLDAP (writes/removes the user-certificates to/from LDAP)
updateLDAP (puts all certs from the last import to LDAP)
(oh, we have a redundancy here updateLDAP and addCertsLDAP do the
same) (updateLDAP is reserved for the future so set all links etc.
to importCertsLDAP) (addCertsUser should not be a function of
ldap-utils.lib
)
LDAP_addCertsUsers (will be moved to importCertsLDAP)
LDAP_get_crl (determines the newest CRL)
LDAP_get_ca (determines the newest CA certificate)
Table of Contents
The most organizations which decide to evaluate or use PKIs make a decision We need a PKI. After some tests, discussions and perhaps some failed roll-outs they start looking for automation tools to limit the chances where end-users can make mistakes which are timeconsuming and in this way expensive. OpenCA tries to solve this problem like many other PKI solutions with a flexible batch system.
The batch system has to fullfill several requirements. It must be
fast
easy to configure
extendable
adaptable to workflows
These are only some requirements. Usually every organization has a bunch of additional problems which should be solved.
The core of OpenCA's batch system is a statemachine which is has primarly nothing to do with OpenCA. This statemachine has a configuration which defines states, functions and some workflows. This statemachine reads at every step the configuration of a user, checks it's actual state, calculates the next function of the workflow which must be executed and returns the function. OpenCA does nothing else then to execute the function for this user. If this sounds really simple to you then please be sure this is only a simplified description :)
First the statemachine doesn't handle users. It handles processes of users. This is necessary because we don't know any installation where every user have only one key and one certificate. At minimum every key must be managed by an own seperate workflow.
Second the statemachine has a two-part configuration. There is a XML file which defines the position of the core configuration files and the options of the batch system functions. The core configuration files are simple text files because they have no structure. They are only lists which are very fast parseable. The core configuration consists of the following parts:
a list of the existing processes (user + key)
a list of all known functions
a list of all known states
a directory which contain the start configuration of the states for every function
Before you now start testing and playing with the batch system one important advice. If you use the batch system then please notice that the default installation uses the CA token as log token too. This means that an activated CA token sign every log entry. This slows down all operations. So if you think that the batch system is too then please remember the small warning about the log token. The issuing of certificates must be done with the log token of course but the time consuming logging should be acceptable in such a critical moment.
Before you can use the batch system you have to import some data. There are three different import actions in OpenCA's batch system:
new users
new processes
data for processes
All three formats are described in the following. You must store
the three files on your dataexchange media in the files
batch_new_user.txt
,
batch_new_process.txt
and
batch_process_data.txt
. After the
specifications you can find an example.
First a description of the formats to create a new user and a new process. Both formats are very simple.
The format to import the data is a little bit more complicated because it must support some special cases.
Please respect the minimum keylength to avoid trouble. __user_defined__ means that you can store other data in the batchprocessor too. This requires that you customize the batch functions. __user_defined__ should be an uppercase word, there are no spaces allowed in the name. This identifier is used as filename, by default we put the file into the "data" directory of the process, if you want to put it somewhere else, you can give the directory after the name with the @ sign, e.g. "PIN@private" will create a file called PIN in the directory private. You can pass mulitline data using mime-style limiters, the limiters will NOT go into the data file, so if you want to pass data which already has limiters and keep them, e.g. a PKCS7 structure, you have to add a second set of limiters. If you need other data which you cannot import with this system then please write a mail and explain your requirements.
Here you find now a very simple example for the two users Jon and Jane Doe.
The above example describes the classic import mechanism with three files. So you can create user, process and data in three single steps, and if you see something going wrong (e.g. creating a user already there) you can stop the process. OpenCA provides a QuickImport Feature that needs only the last file (batch_process_data.txt) and creates user, process and process data in a single step. It will take care and not override already existing process, but create more than one process even if the user exisits in the system. Nevertheless the QuickImport Feature should fit 98% users needs.
First you have to understand that we implemented something called a finite statemachine. If you don't know what this is then take good books on theoretical computer science or ask some professors. Second you have to understand our implementation :) Ok, you have only to understand the configuration. A finite statemachine consists of states and state transition functions (German: Zustandsuebergangsfunktionen).
All known states are stored in
OPENCADIR/etc/bp/states.txt
. You can change
this filename like every other filename too in
OPENCADIR/etc/bp/bp.xml
. All known functions
are stored in OPENCADIR/etc/bp/functions.txt
.
The startconfiguration of states which causing a function to start
it's actions are stored in the directory
OPENCADIR/etc/bp/functions
. The filenames
are function_name.txt
. Every line contains
one state which is required to start the function. Every process
of a user has a configurationfile states.txt
which represents the actual state of the user. It has the same
format like the configurationfiles of the functions.
If you want to change the default workflow or create a complete new workflow then you have to do the following:
Visualize the state-transition graph.
Mark the changes in the graph.
Create OPENCADIR/etc/bp/states.txt
which
must include all possible states.
Create OPENCADIR/etc/bp/functions.txt
which includes all available batch functions.
Create the start configuration of the states for all batch
functions in OPENCADIR/etc/bp/functions
.
Check that the start configuration specified in
OPENCADIR/etc/bp/functions
are conflict
free.
Please perform the last step twice if you change an already running system to ensure that you don't remove a state which is already present in the system and that you are able to still handle all states.
Implement the new and change the old functions.
Test the new system before deploying it.
The implementation of a new function is not the simplest task but it is not very difficult. The biggest problem is the impact that a function can create if it is executed in a not expected situation. If it do something wrong then such a function do it usually wrong for every user. The steps to create a new function are the following ones:
Add the function to the list in
OPENCADIR/etc/bp/functions.txt
.
Create the file
OPENCADIR/etc/bp/functions/function_name.txt
and specify the start configuration of the states there.
Create the OPENCADIR/lib/bp/function_name.sub
by copy and paste from another batch function.
Change the function name in the file
function_name.sub
to
workflow_function_name.
Change the source code of the function to do what you want.
The function get two parameters - USER
and WORKFLOW
. You can use these
parameters to access the data in the batch system.
Finally you have to change the actual state configuration of the workflow.
Please allways log what you are doing. It is the only chance to discover the source of problems. Allways write an entry to the log message if you discover something unusual in a workflow.
Table 16.1. Default OpenCA workflow
Function | Start States | Result States on Success | Result States on Failure |
---|---|---|---|
create_pin | new_process | new_pin | |
check_pin | new_pin | checked_pin | |
check_key_params | checked_pin | checked_key_params | |
create_key | checked_key_params | new_key | |
check_key | new_key | checked_key | |
backup_key | checked_key | backuped_key | |
check_csr_params | backuped_key | checked_csr_params | |
create_csr | checked_csr_params | new_csr | |
complete_csr | new_csr | completed_csr | |
check_csr | completed_csr | checked_csr | |
create_cert | checked_csr | new_cert | |
enroll_pin | new_cert | enrolled_pin | |
enroll_pkcs12 | enrolled_pin | enrolled_pkcs12 |
First you must have a fundamental understanding about OpenCA's concept of crypto tokens. You can find these fundamentals at Chapter 10, Cryptolayer. Second the batch system uses up to three different crypto tokens actively and one passively. The three actively used tokens are the CA, the key backup and the batch token (BP token). The CA token is only used to sign certificates. The key backup token is only used to encrypt a keybackup. The batch token is used for all other things like encrypting PINs and private keys. The passive token is the log token which is allways active or never.
All token must be explicitly activated by entering a valid passphrase before the batch system starts it's operation. The important thing is that by default all tokens are identical. All keys are symbolic links to the CA key. Usually you can activate the batch token and all thinks are ok except that you want to use keybackup or certificate creation. If you enter all passphrases then all tokens are active and you have nothing todo - until the complete system crashs.
1.000 users, each with one process key_1. My machine swapped sometimes because KDE3, Mozilla, PostgreSQL, Apache and OpenOffice is too much for my old notebook.
Table 16.2. 1000 User test
Function | Time | CA-Token |
---|---|---|
total | :: h | - |
create_pin | 2:39 min | off |
check_pin | 30:48 min | off (computer swapped) |
check_key_params | 1:03 min | off |
create_key | 52:39 min | off |
check_key | 32:51 min | off (computer swapped) |
backup_key | 31:47 min | off |
check_csr_params | 1:12 min | off |
create_csr | min | off |
complete_csr | min | off |
check_csr | min | off |
create_cert | min | on |
Table of Contents
Copy the directory suse
to the source
directory
Go to the SuSE area
Build the packages
Optionally it is possible to set a RPM specific package number
to all packages generated (the default is 0). To change the
release number for the generated packages change the RELEASE
variable in suse/Makefile
to the desired
value and run make update-spec-release before
starting the package build. This command locally changes
the 'Release:' tags of all spec files for the
packaging process.
The biggest problems during packaging are the completeness and the conflicts. You can test the completeness really easy by testing the software itself. The conflicts you can only test by installing all the packages one by one. Please do this with only a basic installation of SuSE to avoid missing or wrong dependencies especially in openca-common. Please install the packages in the following order:
openca-doc
perl-CGI-Session
perl-XML-Twig
perl-MIME-Tools
perl-Net-Server
perl-ldap
perl-openca-configuration
perl-openca-xml-cache
openca-sv
perl-openca-openssl
perl-openca-crypto
perl-openca-tools
perl-openca-tristatecgi
perl-openca-session
perl-openca-pkcs7
perl-openca-log
perl-openca-ui-html
perl-openca-ac
perl-openca-req
perl-openca-x509
perl-openca-crl
perl-openca-db
perl-openca-dbi
perl-openca-common
openca-web-interface-node
perl-openca-statemachine
openca-web-interface-ca
openca-web-interface-ra
openca-web-interface-pub
openca-web-interface-ldap
openca-scep
openca-web-interface-scep
openca-ocspd
Table of Contents
Table of Contents
Here you can read some historical notes about OpenCA. Please don't believe they are complete or always rational. We only write such notes if we have no new ideas at the moment ;-D
When OpenCA was just a thought in our mind, the need for a complete PKI solution was already perceiving. The problem, in 1998, was that only few products were available in the scenario and the costs associated with this kind of software was indeed considerable. Most of the PKI software available presented non only payment for the software license, but often there were per-certificate costs which prevented many organization to deploy their PKI.
What were the available software then? Not many, indeed most of the implemented software was developed in USA and, because of rigid export regulations, there were problems associated with the adoption of "strong" cryptography outside United States.
Netscape Company was one of the most advanced provider for this kind of software while other providers were pointing their attention on the selling of certification services instead of software capable of managing a PKI. Some other solutions were available from different vendors in Europe too, but the costs problems remained.
Another example of software provider was Secude. The SecuDE, a german security related software company, was selling a crypto layer for generating certificates and maintaining a PKI without limitations on the key-length. Despite of the fact that this software was not imposing weak keys, the company sold licenses based on the number of issued certificates per year.
Let us make a practical example. Let us pretend a local Public Administration, i.e. a small municipality like Modena or one quite big organization, was about to deploy certification services for its citizens. In this scenario it was logical to be prepared for issuing 50,000 - 100,000 certificates (but it should be considered the worst case where every citizen asked for his/her certificate). The number of certificates and the costs related was so high that no one would ever even considered the chance to deploy the services.
Hence the costs problem was one of the most relevant.
All these problems were present not only in PKI managing software, but in crypto enabled applications too. One noticeable example was the popular Netscape browser (and its competitor Internet Explorer), which was available world-wide only with 56-bit capabilities for symmetric crypto and 512-bit for asymmetric crypto. As we underlined, the need of strong cryptography was very high and that is probably the main reason for the availability of products like Fortify that enabled strong cryptography into Netscape.
The Italian scenario presented some points of excellence, though. Two of these, the Modena's municipality and the Turin's Politecnico, were successfully experimenting and developing software based on semi-proprietary solutions (using crypto libraries like the SecuDE one together with ad-hoc developed software). In this highly experimental phases there were some special form of licensing available from different companies (SecuDE was one of the first but in the next few years other research project have been carried out without challenging the real costs because of specific marketing policies by the issuing companies like, for example, Italian Telecom, BNL Multiservizi and others) that were not tied to a per-certificate basis. But these solutions were only temporary and not viable in a real environment with hundreds of thousands of certificates issued and renewed per year.
The natural choice was, then, to look for some freely available software (better if open source so we could modify to fulfill our needs). The only library which was stable enough and available for free was the SSLeay, a collection of libraries and tools for implementing the SSL protocol with some certification capabilities. The SSLeay from Eric Young was already a widely adopted library (indeed nearly all SSL implementing software and SSH solutions on UNiX already were using it) and it was developed under a BSD like license: this was the solution we were looking for although there was much work to do.
The use of Public Key Infrastructure (PKI) deserves special attention, as the most common challenges for eGovernment to overcome were identified as verifying identity on-line and verifying the authenticity of a document. PKI can address both. The consensus seemed to be that the top level of PKI certificates could be provided like a passport or birth certificate: government defines the format and manages the registry and then provides it via public or private sector, whichever works (like physical passports are delivered).
Some special care, thus, should be taken when addressing PKI's issues as it seems PKI technology can find some very important application in the eGovernment sector. We find interesting that some excellent projects have been carried out using OpenCA in the eGovernment field.
Our project deals mainly with informations and informations need to be interoperable to be useful. Standards are needed in such an environment.
The IETF organization cover every aspect related to the Internet Standards. But what is the IETF? The Internet Engineering Task Force (IETF) is a large open international community of network designers, operators, vendors, and researchers concerned with the evolution of the Internet architecture and the smooth operation of the Internet. It is open to any interested individual.
The OpenCA project has been started when PKI's scenario was lacking either on the applications side and on the certificates' management one. We thought that certification technology, although at hand because of the increasing computational computer power and ever decreasing costs, was suffering by lack of open sourced software easing development of digital certificates' related applications and deployment of users acquaintance with digital certificates. The OpenCA Project was, and still it is, a :
The project development has been divided in two main tasks: studying and refining the security scheme that guarantees the best model to be used in a CA and developing software to easily setup and manage the Certification Authority.
One part of the project was the development of a full PKI Open-Sourced software. The software is actually ever growing as new protocols and standards are being issued by the standardization bodies. The users' and developers' community has been enlarging and we can now count on many individuals and organizations directly or indirectly supporting the project.
Although the software development is one important aspect of OpenCA, our objectives were not only technological. We can say that among the project's objectives there were also the exchange of ideas between different realities and the setting up of a working group deploying code and solutions solving practical certification issues.
Five years have passed since we started to work at the project. OpenCA has been constantly growing and the number and quality of active installations are, today, too many to be mentioned here. We have to state, for clarity, that the project is still evolving and because of this we can only imagine what the future of the project will be. Today the project counts a quite high number of developers, mainly from universities, and contributing users.
The basic principles that have been followed throughout these development years:
adherence to the IETF standards - at every step we tried to stick to the RFCs covering the matter from the very beginning. Indeed the project basic structure is derived from one IETF specification.
feedback from users - the project have grown integrating help and keeping in mind objections by users and developers.
openness - we tried to keep our sources as readable as possible thus allowing almost everyone to contribute with code or comments.
interoperability - this is a consequence of being adherent to the standard but not only. Whenever possible we tried to support the majority of the systems today available. For example we introduced the usage of Java applets to support Microsoft Internet Explorer and we develop support for SCEP.
usage of simple programming language (whenever possible) - this follows from the consideration that security is not obscurity and we needed to be as readable as possible. Then we choose for PERL which guaranteed for portability and simplicity.
Thus most of the project's objectives have been either achieved and yet are still to be achieved. This could be quite a cryptic sentence but it is not. The project has achieved its goals because it helped in searching for solutions about certification problems and issues. For example OpenCA has made possible for some municipalities to deploy services to its citizens and other organizations to develop applications for their customers, we solved many practical problems together with users and developers and we helped in making this technology available for almost everyone. We also contributed to the discussion on the pkix working group about some specific issues like certificates' suspension. Still the project has not yet achieved its goals because much work is needed for this technology to be widely deployed and accepted by users. Many standards and protocols are engineered every day and thus the project still needs active development to be kept up-to-dated.
We started the OpenCA project in June 1998. The very original idea was developed by Massimiliano Pala. The first version of OpenCA has been developed in approximately one monthand the project first sources were composed of a single quite long script. When the first version of the software was developed the OpenSSL project was still named SSLeay and it was very different from what it is now: many functions were still buggy and many others were missing.
Installation of the project was based on a very simple Makefile and some scripts to initialize the CA. To quickly install the software you just needed to unpack the archive, cd to the newly created directory, use the 'make install' command: a script was then run to install the base CA software and generate (and eventually sign) the CA certificate.
A series of script was provided to help in installing and configuring most of the project's parts. Although very simple, this solution caused many problem with the users' community because of issues rising from the need of being able to better customize the full package. For example to be able to generate certificates through the full featured process you needed to install the CA web, RA server software and generate RAs' browser importable certificates. The first version of OpenCA was very basic, functions implemented were aimed primarily to issue certificates and CRLs only and the installation method was quite rude: no usage of any configuration utility (i.e. autoconf and automake) and scripts were compatible with bash only.
New releases were adding ever more features to the project and thus the 0.109a version was already including interfaces for either CA, RAServer and Public. From the very beginning of the project and since the release of the first version a great attention from the Internet community was turned on the project.
Our project has been based on the OpenSSL cryptographic tool. The OpenSSL project was named SSLeay and has been developed originally by Eric Young and then it has been taken in charge by Ralf S. Engelschall and recently by Steve Henson. Although still one of the most advanced (if not the only one in 1998) open source crypto library available, we needed many features not available in the toolkit at the time we started OpenCA. The availability of the sources has been proved essential for us because we have been able to add many new features no one have thought about before. This is not a minus for OpenSSL. This is a plus for open source and OpenSSL will be extended day by day by other developers who also need additional things.
All this work had permitted us to have the needed functions to correctly manage certificates and requests by building a web oriented interface acting as a front end to the crypto layer.
First versions of OpenCA were strictly tied to OpenSSL but we knew we had to abstract from the crypto library as much as possible. If we could be able to have a layered structure it could help external developers to support different cryptographic layers and thus incrementing the project's portability. This has been done since version 0.6.0 which had many new features and a much more complex structure then the former releases.
Because of the constant growth of the software developed and the asking for direct support by the community, we choose to move for advanced development tools. Many people have found themselves to work on different aspects of PKIs and in particular on the Project features and structure.
We used CVS because it helped to manage releases and to control the concurrent editing of source files among multiple authors as new developers were admitted to directly access the main source code.
Thus we needed to install it on our main server, which acted as either package download site and as community support site via the installed web server and different applications: a CVS server and a mailing list management tool. Some difficulties were found for the CVS setup and management because of many security issues we had to care about. Anyway these tools let us work in collaboration with many different developers and contributors from all around the world with little administration efforts.
Security is not obscurity. This is a principle we should all keep in mind when approaching the security issue.
We can therefore state that the Open Source choice, especially when dealing with security and issues tied to this subject, is important because it gives the chance to everyone to study and test our ideas and eventually help in designing a better solution. If we take for true that two eyes are better than one, then why not one hundred or one thousand?
Open Source Projects have indeed another one big advantage: contribution can come directly from users and/or developers who need features (or modifies) you simply have not had the time to think about or to implement. We found that Open Sourcing the Project have given a very big help in growing either in its practical implementation and in its visibility.
We found ourselves at the end of the 2000 with the need to have more administration power, to coordinate different developers, and Internet connectivity. We have been asked to move on different servers by the users and proposals were not missing. Anyway we decided to move to SourceForge. At the moment of this writing we have many mirrors around the world updated daily, mainly thanks to the Sunsite and SourceForge networks, giving us much visibility over the whole Internet. This helped especially Asian users that suffered many times from lousy connections and long download times.
The SourceForge was then the natural choice where to move the project core to. This has saved much administration work and thus has let us work much on the project than on the development tools used.
Table of Contents
Many people ask who uses OpenCA. The problem is that we are a non-profit organization and so we have real problems to get references of existing systems. If you, your organization or your company use OpenCA and have no problem to publish this fact then please mail us. This helps us to give others an overview and they can ask a local user for it's experiences. This is sometimes more important than any other fact. We can add a small description too.
We use OpenCA in our lab & have delivered the same to our main university for their internal use. Infact we started giving commercial support to some of our local customers too.
MIT (Madras Institute of Technology), Anna University Campus
Chennai - 44, TN, India
Contact: <mmurali@au-kbc.org>
We use three OpenCA PKIs.
Two Sub-CAs
in the hierarchy of the
DFN-PCA. This is
the Root-CA of
Germany`s National Research and Education Network.
One CA is a private isolated root CA which is based on OpenCA 0.9.1. We
use this for internal communication including key recovery.
You can contact us via <pki@hu-berlin.de>
.
We are running one OpenCA since 2003 and use it for secure eMail for our students and employees (approx. 2500). Currently we set up a new structure based on the 0.9.2 release with one Root-CA and two Sub-CAs for eMail security and keybased shell access. We plan to certify the system within the hierarchy of the DFN-PCA - the PKI of Germany`s National Research and Education Network. See http://www.ldv.ei.tum.de for details - questions are welcome.
OpenCA supports internationalization (i18n) since version 0.9.1. Full i18n support is available with version 0.9.2. There are several different translation teams. Following you can find the available translation and the contact informations. Please notice that we usually don't publish private addresses here to reduce the spam of the authors. If there are real problems then we have the emailaddresses of course and we will provide you with them. If you want to create a new translation feel free to contact us on one of our mailing lists.
The German translation is managed by Michael Bell from Humboldt-University of Berlin but there are more than one German OpenCA developer. So the easiest way is to write a mail to openca-users@lists.sf.net if you have a problem with the translation.
Massimiliano "madwolf" Pala is the actual manager of the Italian translation. He is OpenCA's project manager and the original founder of the project. You can contact him via our mailing lists too.
The Japanese translation was created by Takahiro Tsuji. You
can conact him for any problems at
<openca@poplar.complex.eng.hokudai.ac.jp>
. He works
at the ITS (Information Technology and Systems) laboratory of
Hokkaido University in Japan. It was the first translation which
don't based on conventional characters.
Franciszek "Franz" Lewenda from the Research and Academic Computer Network (NASK) creates the Polish translation. You can reach him via our mailing lists. This was the first non-iso-8859-1 translation.
The Slovene translation was created and is for now managed by Janez Pirc. He works for SETCCE (Security Technology Competence Centre, http://www.setcce.org) which uses OpenCA to manage their own Certificate Authority under the name SI-CA (http://www.si-ca.org). This was the first UTF-8 translation of OpenCA. If you want to contact the translator, just send an e-mail to openca[at]si-ca[dot]org.
Table of Contents
Here you can find all the authors and contributors to this guide.
Configuration issues, batch system, LDAP, FAQ maintenance and all content which is not explicitly written by others :)
Several people have helped us on the mailing lists. We have extracted many hints and workarounds from these lists and have added them to the FAQ. We try to list here the guys that have helped us. If we have forgotten somebody please help us and send a notice.
Dejan Kulpinski - Microsoft Smartcard Logon
Gottfried Scheckenbach - Microsoft Outlook Express
Table of Contents
-subj
of openssl req/dev/fd0
A certificate is a so called digital ID card. The correctness of a certificate will be guarnteed by a certificate from a higher level of the hierarchy. Such a certificate is called CA certificate.
Certificate Informations
serial number of the certificate
a subject (name)
the corresponding public key to the private key of the certificate owner
the name of the issuer
the version of the certificate
the used cryptoalgorithms to create the certificate
the validity period
some extensions
the digital signature of the certificate
There are two types of requests CSRs and CRRs. CSRs are used to ask a trustcenter for a certificate. CRR are used to ask a trustcenter to revoke a certificate if it is corrupted. There are two important standards for CSRs - PKCS#10 and SPKAC. OpenCA can handle both standards automatically.
CSR Informations
a subject (name)
the version of the request
the corresponding public key to the private key of the certificate owner
some attributes
It is fairly well known that there are two versions of Xenroll.dll used by versions of IE to create certificate requests and manage CSPs etc.. OpenCA since version 0.9.1 has managed them via the ieCSR.vbs scripts.
We have noticed that if a user has Win2K and IE6 SP1 then the version of xenroll.dll does not work and the users can see no CSPs to manage their certs with. A patch is required from Microsoft (323172) for Win2k, or it needs to go up to SP3. You can host a copy of the latest xenroll.dll on your web site under a CertControl directory and it will be downloaded and installed automatically.
As far as we can tell, the latest xenroll.dll is a different file, but shares the same identifiers as the pre-patched version. We have noticed that the isCSR.vbs (as of 0.9.1) is written in such a way as to not expect there to be a non working version of xenroll.dll, so there is a bit of a gap.
Initialize the SubCA (initialize database, generate secret key, generate request)
export request
untar the export (to get the careq.pem), the next steps are only correct if you use OpenCA for your Root CA
Point to the Root CA public interface -> request a certificate -> server request -> browse for the careq.pem and submit the request
Point to the Root CA RA interface and approuve the request, upload to the Root CA CA; point to CA interface, issue the certificate
Download the certificate for the sub CA via the RA or public interface of the Root CA
rename the file to cacert.pem and manually make a new tar
Point your browser to the SubCA CA interface and import CA certificate approuved by Root CA
Yes, it is possible. Go to a RA interface. Go to the certificate which you want to revoke. View the certificate. Click on revoke, fill out the form and now you have created the initial CRR to revoke the certificate.
This message appears if one of the configurationfiles of the new
role already exist. Please check the files in the directories
OPENCADIR/etc/openssl/extfiles
and
OPENCADIR/etc/openssl/openssl
.
Check that the configuration option OPENSSL
is
set to the correct path. It mus be the binary of OpenSSL. You
have to verify all files in
OPENCADIR/etc/servers/
.
You are using OpenSSL 0.9.6 but you must use 0.9.7. The use of
0.9.6 can cause inconsistent data. Normally OpenCA cannot
installed if OpenSSL 0.9.7 is not present. So please check the
path to the OpenSSL binary in the configuration files. The
option is OPENSSL
in all files in
OPENCADIR/etc/servers/
.
Please check the settings in
etc/servers/DBI.conf
because this happens
if IBM's software cannot find the libraries and databases.
it is now possible to create usable packages
you can configure the PKI after the installation
docbook based documentation
integrated access control
secure export of private keys via the public interface
several LDAP improvements
keysizes are now choosable for IE users too
much better CSR editing
additional attributes for requests (e.g. telephonenumbers)
menugeneration via XML-configurationfile
SCEP support
warn expiring certificates
more (an explicit) download formats for certificates
subject verification for PKCS#10 requests
logging support
Mozilla doesn't implement crypto.signForm until version 1.7. We strongly recommend that you update to a newer version. Some workarounds are described at Section 2.2.2, “Signing Data”.
KDE doesn't include any functionality to sign HTML forms until know. So this feature is not supported for KDE.
It is a noncompressed tar file. The name of the file which contains the CA certificate is cacert.pem. The format of the file is PEM (sometimes called CRT or base64 encoded).
If you try to create a CRL, to issue a certificate or to revoke
a certificate and it fails then you should get an errormessage
from OpenSSL. If the errormessage include the string
entry 1: invalid expiry date then the
database file index.txt
is damaged. The
easiest solution is to go to the backup and recovery are of the
node management interface. There you can use the link which starts
the rebuilding of the OpenSSL files. After this operation the
OpenSSL files are correct again.
If you imported the certificate of another user and try to send him an encrypted email then it can happen that this doesn't work with Outlook and Outlook Express. The reason is that the person must be present in your contacts. The best way to add the person to your contacts is to take a signed email and import the user from this email to your contacts.
There are several events why Outlook freezes but one events is a signed email in combination with an anti virus program. One user reports some time ago a frozen Outlook in combination with an anti virus program from Kapersky. Like often with Microsoft programs it is not clear why Outlook crashs and who makes the mistake and includes a bug in it's program.
If you try to issue a certificate and you use an OpenCA version prior to 0.9.2 then it is possible that you get a general error 6751.
Example E.1. General error 6751 during certificate issueing
Error 6751 General Error. Error while issuing Certificate to CA Services some.host.com (filename: /usr/local/openca/var/tmp/04.req). OpenCA::OpenSSL returns errocode 7731071 (OpenCA::OpenSSL->issueCert: OpenSSL fails (256).)..
If you check your Apache's error_log
then
should see some lines which include digital envelope
routines:EVP_DecryptFinal:bad.
Example E.2. Bad passphrase error log during certificate issueing
[Mon Dec 29 18:32:59 2003] [error] [client 192.168.1.38] unable to load CA private key, referer: http://ca.localhosts.com/cgi-bin/ca/ca?cmd=viewCSR;dataType=APPROVED_REQUEST;key=1312 [Mon Dec 29 18:32:59 2003] [error] [client 192.168.1.38] 18685:error:06065064:digital envelope routines: EVP_DecryptFinal:bad decrypt:evp_enc.c:438:, referer: http://ca.localhosts.com/cgi-bin/ca/ca?cmd=viewCSR;dataType=APPROVED_REQUEST;key=1312 [Mon Dec 29 18:32:59 2003] [error] [client 192.168.1.38] 18685:error:0906A065:PEM routines: PEM_do_header:bad decrypt:pem_lib.c:421:, referer: http://ca.localhosts.com/cgi-bin/ca/ca?cmd=viewCSR;dataType=APPROVED_REQUEST;key=1312
The reason is very simple. The messages unable to load CA private key and EVP_DecryptFinal:bad decrypt are from OpenSSL and signal that the CA's private key cannot be decrypted. This usually happens if you use a wrong passphrase. You can test your passphrase with the command openssl rsa -in /usr/local/openca/var/crypto/keys/cakey.pem -text -noout. If it fails then your passphrase is wrong.
This defines all necessary steps for a new release and is mandatory for release candidates too. Steps which are on mandatory for normal releases or release candidates are marked.
Go to CVS module directory openca-0.9
Edit Makefile.devel
and fix the minor release
Commit Makefile.devel
cd ..
cvs tag -R openca_V_E_R_S_I_O_N openca-0.9
cd openca-0.9
make -f Makefile.devel dist
scp openca-0.9.2*.tar.gz username@ftp.openca.org:ftp/releases/
ftp upload.sf.net
Login: anonymous
Passwd: your emailaddress
cd incoming
put openca-0.9.2*.tar.gz
Go to sourceforge.net and release the file for project openca
Add a release for OpenCA at freshmeat.net
Add news message to news area of OpenCA.org
Send a mail to openca-users, openca-devel, openca-announce
Which string should be filled in the Service URL
field of the Mozilla Preferences/Validation
assuming that 10.13.1.13
is my CRL IP?
Well, it depends on your configuration (check the
ocspd.conf
). Anyway by default you should
use http://10.13.1.13:2560/
.
Sometimes you get the following error message.
Example E.3. Error 7211021: Cannot create request!
Error 7211021 General Error. Cannot create request! (OpenCA::REQ->new: Cannot create new request. Backend fails with errorcode 7712071. OpenCA::OpenSSL->genReq: Cannot execute command (7777067). problems making Certificate Request 24649: error:0D07A097: asn1 encoding routines: ASN1_mbstring_copy: string too long: a_mbstr.c:154:maxsize=2 error in req ).
The reason is very simple you entered more than two characters for the ISO country code. Please check you form and the configuration for the used country code. All ISO country codes are two characters long - not one character and not more than two characters.
Please ignore this section for OpenCA 0.9.2.0+ and snapshots after 2004-Sep-15.
FreeBSD and potentially several other BSDs use an own make. Nearly the complete development of OpenCA is done on Linux. The result is that our makefiles works perfectly with GNU make but there are several unresolved problems with FreeBSD because we cannot test and fix on FreeBSD. The solution is to add GNU make to configure. You can do this this way:
Please check Makefile.global_vars
to include the
correct setting of MAKE
. If the correct setting of make
is missing then please add the following line:
GNU make is fully tested with OpenCA.
There is an additional problem with install
like with make. FreeBSD doesn't support the
parameter -D
. This parameter is only required if
some of the used installation paths like BINDIR
do not exist. If you perform a normal installation and use default paths
then there should be no problems without -D
.
You have to check Makefile.global_vars.in
before
you run ./configure or
Makefile.global_vars
after you run
./configure. There you have to remove the
parameter -D
from the definition of the variable
INSTALL
.
If you are not the hacker which like to dig for bugs and the removal of
-D
doesn't work then you have one final option.
You can install the port "coreutils" which includes the GNU install.
this is not a real bugfix but it works.
Please ignore this section for OpenCA 0.9.2.0+ and snapshots after 2004-Sep-15.
Solaris has the same problems like the BSDs. The Solaris make doesn't work with OpenCA's makefiles. The solution is to install the binutils and gmake from sunsolve. You can configure OpenCA after the installation at this way:
Please check Makefile.global_vars
to include the
correct setting of MAKE
. If the correct setting of make
is missing then please add the following line:
GNU make is fully tested with OpenCA.
OpenCA 0.9.1 and earlier includes some protection mechanism to
avoid state injections during the dataexchange. Therefore you
have to configure the level at which OpenCA performs the
dataexchange, so that ./configure can make a
good preconfiguration. This is only correct for 0.9.1. 0.9.2+
uses OPENCADIR/etc/config.xml
to configure
the dataexchange.
If you export from a public area to a server with the RA
then must use pub
.
If you export from a server with a RA to a server with a
CA then you have to use ra
.
If you export from a server with a RA to a server with a
public interface then you have to use
ra
.
If you export from a server with a CA to a server with a
RA interface then you have to use
ca
.
If you have another case then you have to chosse one of the above defined options and then you have to edit the options for the dataexchange manually like described in the administrator guide.
Usually the errortext is a little bit more descriptive and looks like this:
Example E.4. Full errormessage for missing functions
[Thu Oct 09 14:50:52 2003] [error] [client 127.0.0.1] Undefined subroutine &main::configError called at /var/www/cgi-bin/ca/ca line 86., referer: http://localhost/htdocs/ca/
The reason for such an error is a missing library - in this case
misc-utils.lib
. There is only one
well-known error - you used
--enable-package-build
as configure option.
This happens if somebody uses a configure example which is used
for package builds. The disables the installation of all common
software parts like modules, libraries and configuration files.
If you configure again without
--enable-package-build
and then run
make install-xyz again then all
libraries should be present.
Several user observed that OpenCA 0.9.1.x installation failed during the installation of symbolic links for commands. The errormessage simply reports an already present symbolic link.
Example E.5. Already present symbolic link
make[8]: Entering directory `/home/linux/tar/openca-0.9.1.2/src/web-interfaces/ca/cmds' cd /usr/local/openca.0.9.2/openca/lib/servers/ca/cmds; \ ln -s ../../../cmds/add_module ln: `./add_module': File exists make[8]: *** [/usr/local/openca.0.9.2/openca/lib/cmds/add_module] Error 1 make[8]: Leaving directory ... make[1]: Leaving directory `/home/linux/tar/openca-0.9.1.2' make: *** [install-ca] Error 2
The reason for this failure is a defect symbolic link. The makefile removes an already present symbolic link only if the symbolic link is correct. If the target of the symbolic link doesn't exist then the symbolic link is not removed and the error occurs.
If the command make install-* doesn't report
an error and after the installation all common parts like modules,
configuration, libraries, openca-sv and commands are missing then
please check your configure options. Sometimes the people simply
copying our configuration examples without noticing that some of
these examples are used for package creation. These examples
include an option --enable-package-build
. This
option prevents all common stuff from installation to build
better packages.
Sometimes there are reports about errormessages which signal an unknown login type or another empty configuration option. First you should check that there are at minimum two running processes - one is the OpenCA daemon which handles the requests and one is the XML cache. If there is only one server then there is a problem during the startup of the daemons. There two well known issues during the startup - wrong permissions and duplicate installations of a module. Please allways check first that there are no problems with the permissions.
If you are sure that you have no problem with the permissions then it is usually a problem with different installed version of the Perlmodule XML::Twig. This can cause a crash of the daemon at startup or always empty results for a configurationfile. First you should search for the different version of XML::Twig.
Example E.6. Search for XML::Twig modules
## every Unix find / -name Twig.pm -print ## rpm based Unix (RedHat, SuSE, Mandrake ...) rpm -qa | grep -i twig ## dpkg based Unix (Debian) dpkg -l "*twig*"
After you know the installed modules you have to remove all versions older than 3.09. OpenCA was developed with 3.08 and 3.09 but the last tests were only performed with 3.09. So we don't know what's happen if you use an earlier version.
If you installed OpenCA and try to login to OpenCA then it can
happen that you see the errormessage The xml path to
the access control is missing. The errorcode should be
6292010. Usually the missing XML path is
access_control/acl_config/module_id
.
Please check that a XML parser is installed on your system. We use XML::Twig to parse huge XML files. This module uses an already installed XML parser. If there is no XML parser installed for XML::Twig then we cannot read and interprete our configuration files.
Sometimes you get the errormessage An unknown login type was specified in the configuration! after you installed OpenCA and try to login for the first time. The errorcode in this situation should be 6273966.
Please check the UID of the socket which the XML cache uses. Sometimes the UID of the socket is root and not the UID of the webserver. We don't known how this can happen. Please verify that this socket is not owned by root.
To solve the problem you can stop openca with
openca_stop or
openca_rc stop.
Then you have to remove the socket
OPENCADIR/var/tmp/openca_xml_cache
.
After this you can start openca again and please verify that the
permissions and owner are now correct.
If an Internet Explorer reports a type mismatch during the request
generation then please check the configuration of your Apache.
Usually the reference to the VB script ieCSR.vbs
is wrong. This happens if the document root of the Apache is wrong.
Example E.8. Failed startup with wrong Net::Server version
#> /etc/rc.d/init.d/openca start Starting OpenCA: Process Backgrounded 2004/04/22-12:34:19 OpenCA::Server (type Net::Server::Fork) starting! pid(3195) Binding to Unix socket file /var/lib/openca/tmp/openca_socket using SOCK_STREAM Setting gid to "73 73" Setting uid to "73" Couldn't POSIX::setuid to "73" []
This errormessage is caused by an outdated perl module Net::Server. Please update to a Net::Server version higher or equal to 0.86. The server itself will usually starting nevertheless the error is reported or not. Please use ps -efa to check for the two necessary daemons - XML cache and OpenCA server.
If you see one of the following errormessages during the compilation or installation of OpenCA then you are missing the perl module XML::Parser.
The first symptom can be a failing make in XML::Twig. Please notice that a second make in the same directory does not show any anomaly. Sometimes only the first make shows up an error message because of the missing XML parser.
The second symptom can happen if you try to configure your OpenCA software.
If you try to locate the installed Twig.pm
files and check there length then usually only
Twig.pm.slow
is not empty.
Here is a small example for the configuration of a virtual host for OpenCA.
Example E.12. virtual host configuration
<VirtualHost _default_:443> ServerName 157.159.100.42:443 ServerAdmin pascal.verrecchia@int-evry.fr DocumentRoot /srv/ra/apache/htdocs ErrorLog /usr/local/apache/logs/error_log Options MultiViews Indexes Fol ................ ........... </VirtualHost>
It is important to bind the address and the port this mean that you should include the following statement in you httpd.conf:
BindAddress Your_address_IP:80 Listen 80
You can also add BindAddress
*
to be sure.
Here is a small example from an OpenCA user which the developers never planned but it works.
Example E.13. ./configure and virtual hosts
--with-ca-htdocs-url-prefix=http://ca.dskt6807.zhwin.ch \ --with-node-htdocs-url-prefix=http://node.dskt6807.zhwin.ch \ --with-ra-htdocs-url-prefix=http://ra.dskt6807.zhwin.ch \ --with-ldap-htdocs-url-prefix=http://ldap.dskt6807.zhwin.ch \ --with-pub-htdocs-url-prefix=http://pub.dskt6807.zhwin.ch \
The example is from OpenCA 0.9.1. Please do these configuration
in config.xml
if you use OpenCA 0.9.2 or
later.
Yes, it is possible. There is an option
LDAPexcludedRoles
in the configuration files of
the node and the ldap interface. If you add a role there then
all certificates which have this role will not be published via
the LDAP server.
Yes, you can edit the httpd.conf in the appropriate way. Please don't do this for the public interface.
Example E.14. Client authentication with mod_ssl
<VirtualHost ra.mycompany.de:4443> ServerName ra.mycompany.de DocumentRoot /RA/apache/htdocs ServerAdmin nicolaie.szabadkai@mycompany.de SetEnvIf User-Agent ".*MSIE.*" nokeepalive ssl-unclean-shutdown downgrade-1.0 force-response-1.0 SSLEngine on SSLCertificateFile /RA/ssl.crt/server.pem SSLCertificateKeyFile /RA/ssl.key/key.pem SSLCertificateChainFile /RA/OpenCA/var/crypto/chain/cacert.crt SSLCACertificateFile /RA/OpenCA/var/crypto/cacerts/cacert.pem SSLCARevocationFile /RA/OpenCA/var/crypto/crls/cacrl.pem SSLVerifyClient require SSLVerifyDepth 10 SSLOptions +StdEnvVars +ExportCertData +StrictRequire ErrorLog /var/log/httpd/ra.srv.err.log CustomLog /var/log/httpd/ra.srv.req.log "%t %h %{SSL_PROTOCOL}x %{SSL_CIPHER}x \"%r\" %b" ScriptAlias "/cgi-bin/" "/RA/apache/cgi-bin/" <Directory "/RA/apache/cgi-bin"> AllowOverride None Options FollowSymLinks Order deny,allow Deny from all Allow from 10.1.114 10.100.1 10.1.102 SSLRequireSSL SSLRequire ( %{SSL_CLIENT_S_DN_O} eq "MyCompany" \ && %{SSL_CLIENT_S_DN_CN} =~ m/ramanager?/ ) </Directory> ... </VirtualHost>
If you no go to the RA then you have to choose a certificate which your browser will use to authenticate you to the Apache. Please note, that for Apache 2.0.* you also need to add in the line "SSLOptions +OptRenegotiate" in order for Apache to re-negotiate the access permissions correctly.
It is necessary that you update your OpenCA installation from time
to time. This can happen if there is a new security advisory or
some normal bugs are fixed.
Since OpenCA 0.9.1 there are bugfix releases. These
releases only update the software. They never touch the
configuration. If you have a 0.9.1.3 and you have to update to
0.9.1.4 then simply use the same configuration like for 0.9.1.3.
The installation with make never
overwrites the etc
or
var
area of your already existing
installation. Nevertheless it is strongly recommended to backup
your complete installation before you start such critical
operations like an update.
There are two general methods. You can use backup/restore or you add the columns by hand.
If you want to use the first possibility then you have to do the following:
Create a backup from the database with the node interface.
Remove all tables and sequence generators.
Update OpenCA.
Go to the node interface.
Initialize the database.
Go to the recovery page of OpenCA.
Import the database.
Restore OpenSSL's files.
If you want to update the database by hand then you have to do the following:
Login to the database.
alter table request add scep_tid TEXTTYPE; alter table log add scep_tid TEXTTYPE; alter table request add loa TEXTTYPE; alter table certificate add loa TEXTTYPE; alter table crr add loa TEXTTYPE; alter table log add loa TEXTTYPE;
TEXTTYPE differs for every database. The following table contains the correct type names for every supported database.
Table E.1. Texttypes for different databases
Database | type |
---|---|
mysql | TEXT |
Pg | text |
DB2 | long varchar |
Oracle | varchar2 (1999) |
Go to the node interface.
Administration.
Databasehandling.
Update searchable attributes.
Some releases include an incomplete sample config. You have to add something like this to your ocspd.conf:
Example E.15. OCSP configuration for LDAP
[ OCSPD_default ] .... dbms = ocsp_crl [ ocsp_crl ] crl_url = ldap://my.ldap.server crl_entry_dn = cn=MyCA,ou=CA,o=MyOrg,c=MyCountry
Alternatively you can use http too:
Example E.16. OCSP configuration for http
[ ocsp_crl ] crl_url = http://my.ca-public.server crl_entry_dn = /crl/cacrl.crl
Before you run configure with the changed config.xml for the second public interface you have to reduce the scope of the files in configure_etc.sh to the new interface.
After such a crash you can configure
config.xml
to the old values, set the paths
in configure_etc.sh
to the first interface
only (!!!) and then run configure_etc.sh
again.
Please read the notices about SMTP servers in the OpenSSL section of the administration guide. If you only have one certificate for your mailserver then it must include the extensions for SSL servers and SSL clients. The extensions for SSL servers are not enough because SMTP servers act as clients too.
There are some situations where clients hang after they try to connect to a TLS or SSL secured server. Examples are Microsoft Outlook clients which connect to mail servers which use TLS or Microsoft Internet Explorers which try to connect to a https server.
Usually the certificate contains a CRL distribution point (CDP) which uses https or ldaps as protocol. The result is that the client tries to verify the server certificate and opens a connection to the server which stores the CDP. If this server presents a certificate which contains a CDP with TLS protection then you have a perfect loop. This can also happen if you try to verify a client certificate which includes a TLS or SSL secured CRL distribution point.
There are two solution for this problem. First you can use only http and ldap or other supported protocols for CRL distribution which don't use TLS and SSL. This is not a big security risc because CRLs are protected by the signature of the corresponding CA. Second you use https or ldaps for client certificates but http or ldap for server certificates. This will produce only one loop if the server certificate will be verified.
If you already enrolled an infrastructure and now you are running into problems with hunderds or thousands of client certificates then you should use the second option to solve your problems. If you enroll new certificates for the servers then you have no problems with your endusers - you have not to explain the problems, the installation of new certificates and the reasons why you don't expect such problems. You “only” install some new server certificates and all problems are fixed like a simple network problem.
The CDP of the certificate from the signature points to a SSL-secured website which was signed by the same CA than the mail certificate. Best solution: Change the CDP to non-https url or a https-url signed by another CA and reissue the mail certificates. If you dont want to reissue all your mail certs it's ok to just change the webservers CDP URL and reissue the webserver certificate.
Old versions of OpenCA include a hardwired minimum length for
HTML-textfields. The minimum length was three. You can change
this limit in basic_csr
. New versions of
OpenCA can be configured. Please read the “installation
and configuration guide”
Section 4, “CSRs”.
Yes, please check etc/config.xml
. There
are two options menu_logo_left
and
menu_logo_right
which can be used to place
logos in the menuframe. Please be careful with this feature
because it can reduce the usability of the software.
The correct errormessage is usually:
Please check the configurationfile
OPENCADIR/etc/token.xml
.
The reference to OpenSSL must be correct with the full path and
not only openssl
.
You have to edit two files - OpenCA's ca.conf
and Chrysalis-ITS's Chrystoki.conf
. First you
should configure OpenCA in
OPENCADIR/etc/servers/ca.conf
.
After this you must configure the Luna device.
Now OpenCA 0.9.1 should be ready for Chrysalis-ITS LunaCA 3.
If you enroll a certificate and a private key to a user via file in PKCS#12 format then you usually want to include the complete certificate chain. This is necessary because many software products doesn't work if the chain is incomplete. This can be normal mail programs or VPN clients. The price is no argument in this case.
Otherwise there can be problems if you try to install certificates which are already present at the target system. The worst case is the destruction of already exeisting certificate chains by overwriting an old CA certificate. Therefore OpenCA only includes the CA certificate which issued the enrolled certificate. Nevertheless it is possible to include as many certificate as you want.
Here are the steps to include other certificates into the PKCS#12 file - it is a typical Open Source solution:
Go to OPENCADIR/lib/cmds
Edit send_cert_key_pkcs12
. There is a line
my $cacert = getRequired ('....
This line defines a file which includes all (CA) certificates which will be included into the PKCS#12 file. Usually we only include our CA certificate. Now you have to setup an individual for your chain.
my $cacert = "/my/openca/dir/var/crypto/cacerts/blaine.pem";
The next step is to create an individual file for the chain.
Now you have to create the file blaine.pem
.
This file has to include all needed CA certs in PEM format.
Please remember to include a begin and end line before and after
every CA certificate like for every normal PEM-formatted CA
certificate.
If you get an errormessage which indicates that there was an unknown login configured then please see Section 3.7, “Conflicting Modules”.
If you get an error message that OpenCA cannot initialize the
cryptoshell and after the startup no daemon is up - wether the XML
cache nor the OpenCA daemon then you have perhaps uncleanly
stopped OpenCA's daemons. Please check
OPENCADIR/var/tmp/
for any socket files. If
OpenCA was killed during shutdown without an explicit shutdown by
openca_rc
or openca_stop
then the still existing socketfiles can block the startup of the
daemons. This behaviour is a must to avoid implicit killed daemons
if you try to start your daemon twice.
How can one create a CA certificate so that the DN does not
contain the email address, but the alternate name does? You
can configure this in
OPENCADIR/etc/openssl/openssl.cnf
.
Set the subject alternative name for the v3_ca to the required
emailaddress. Don"t add the emailaddress to the subject of the
new request.
Sometimes users see the following error message:
This means originaly that you are using a symmetric key which is
shorter than specified in
OPENCADIR/etc/access_control/ra.xml
. Usually
the symmetric cipher must have a length greater or equal 128. If
you are using a Mozilla browser then you can klick on the small
lock to get informations about the used session cipher.
The empty brackets at the end of the errormessage means that there is no keylength specified. This is a perfect indication of a problem with the environment variables. Usually the standard environment variables for SSL are not activated in your Apache. You can confgure it like follows:
Please NEVER set the required symmetric keylength to 0. This is security hole.
Sometimes there are problems with PKCS#10 requests. OpenCA displays an error messages like this one:
Please check config.xml. Usually you forgot to configure the
country name. If you don't want to run
configure_etc.sh again then please check the
files in OPENCADIR/etc/servers/
for the
configuration of the country code.
You try to login with correct login and password. The result are one or two frames with the login screen again.
If your credentials are really correct then there is in the most cases a problem with the Perl module CGI::Session. Old versions of this module have problems with the flushing of data- especially the automatic flushing of data. OpenCA tries to handle this bug but the best solution is to remove the original module from your computer and to install a new one. You can install a new version from CPAN or you install OpenCA again. If OpenCA cannot find this module then it installs an actual version by itself.
This error message happens if you are using an apache without
mod_ssl but you specified mod_ssl
in
your access control configuration.
This happens if the configuration requires
ssl
but you are using http to access the interface.
The access to the interface is only allowed for some computers with special IP addresses and you have not a computer with one of these special addresses.
There is a problem with the algorithms which mod_ssl and your browser are using. The configuration tries to enforce some special algorithms for security reasons. Please contact your administrators for more informations.
The keylength of your certificate is too short if you are using a certificate. If you don't use a certificate then the Apache itself or your browser only support some too short asymmetric keys. Please contact your local administrators to find out more details.
There is a problem with the algorithms which mod_ssl and your browser are using. The configuration tries to enforce some special algorithms for security reasons. Please contact your administrators for more informations.
If your browser startis a connection to webserver via https then the two software components try to negotiate about the cryptoalgorithms which they support. Usually they choose the strongest algorithm and longest keylength. Some export restricted browsers only support too short keylength. This is not allowed in security sensitive areas. Please ask your administrators for the exact specifications of your installation.
This is the most common problem with the export of data. The most modern Unix systems change the owner of local drives (if you can change the medias) to the UID and GID of the user who is logged in on the console. If you want to export something to the floppy then there are three choices:
The simplest solution is to change the permissions itself and to allow every other user to write on the disk.
chmod a+w /dev/fd0
If you run no X server or better no xdm (including gdm
and kdm etc.) then it makes sense to simply change the
group and perhaps the owner of the device. This
operation makes no sense if you run xdm because there
is a PAM module for device settings and this PAM module
is usually included into
/etc/pam.d/xdm
.
chown wwwuser:wwwowner /dev/fd0 chmod u+w /dev/fd0
If you run OpenCA on a production system and you run an xdm on this system then you should use this way of changing the permissions but be prepared this is not very easy and you must test the changes very carefully.
First you have to edit
/etc/logindevperm
. Please
comment out the line which defines the settings for
/dev/fd0
. This will avoid the
PAM module devperm from
changing the owner (UID) and group (GID) of the
device. Usually the PAM module is used by xdm
(see /etc/pam.d/xdm
) and other
console based services.
Now you can do the same like for the first or the second choice.
chown wwwuser:wwwowner /dev/fd0 chmod u+w /dev/fd0 or chmod a+w /dev/fd0
Now the change of the permissions is durable. The changes in the first and second choice are not durable if you use the PAM module devperm. Please don't wonder if a normal user cannot mount a floppy after these operations.
Do you see the following during the import:
Test the archive ... /bin/tar -tvf /dev/fd0 Importing archive ... Load required variables ... Changing to directory /usr/local/openca.0.9.1/openca/var/tmp/tmp_418 ... Running the import command(s) ... /bin/tar -xvf /dev/fd0 -C /usr/local/openca.0.9.1/openca/var/tmp/tmp_418 Importing the RBAC-configuration ... Ok. LDAP-support is activated Automatic LDAP-update is activated Importing _CA_CERTIFICATE ... No objects are present. Importing CA-Certificates into ldap ... Cannot load CA-certificate Make CA-Certificate available on the server ...OK. Re-Building CA Chain ... Ok. Clean up ...Ok.
The important thing is the line:
Importing _CA_CERTIFICATE ...
If you see this line then your configuration for the
dataexchange is wrong. This can happen if you installed the
online components with
--with-hierarchy-level
=ca
.
A CA doesn't import of course a CA certificate.
You can check the option
DOWNLOAD_CA_CERTIFICATE_STATES
in your
configuration files. It should contain at minimum
VALID
.
Go to OPENCADIR/var/log/enroll
and
OPENCADIR/var/log/download
. There you
have to remove all data from files which contain the module ID
of the node interface on the online server which crashed. If you
run the import/export commands next time then all objects will
be exported again. It is the same technology like for the
creation of a new node interface.
Usually the log from the export shows no requests or looks like this:
Example E.21. Failed request upload
Exporting _REQUEST ... FAILURE: 288 (spkac). FILE: /srv/ra//OpenCA/var/tmp/tmp_1517/_REQUEST//288.spkac FAILURE: 544 (spkac). FILE: /srv/ra//OpenCA/var/tmp/tmp_1517/_REQUEST//544.spkac Exporting archive ...
The symptom is “Exporting _REQUEST”. This shows that there are no status specified for the export of requests. New version of OpenCA simply ignore the requests and show nothing.
You can manually change the status of the requests which
should be uploaded or you install again. In the most cases there
was a wrong paramter for configure. Please check the value of
--with-hierarchy-level
very carefully. If you
install a RA and you want to export approved requests to the CA
then the hierarchy level must be ra
.
This occurs if OpenCA cannot make a connection to the LDAP
directory. Make sure that the ldap server is running and is
listening on the correct port. Make sure that the settings in
ldap.xml
match your ldap server settings.
A connection has been made to the ldap server, but the
credentials to log into the server as admin are wrong. The bind
operation is performed after the connection is established.
Check the login
(the LDAP administrator's DN)
and passwd
(the password of the ldap
administrator).
This sometimes means that OpenCA could not insert the
appropriate entry for a certificate (the exact definition is
LDAP_OBJECT_CLASS_VIOLATION
). Check that
you have the directory started with the appropriate schemas
(core, cosine, inetorperson and openca). They are usually
specified in slapd.conf
.
You can get more debugging informations by turning on debugging
in OPENCADIR/etc/ldap.xml
(i.e. <debug>1</debug>
). The
most functions support this paramter.
The logging messages of OpenLDAP are sent to syslogd. OpenLDAP
uses the facility local4. You can find the files which contain
the logs in /etc/syslog.conf
. Simply search
for the files which will be used by local4.
If you need more informations than be in the log files from
syslogd then you have to tune the configuration of OpenLDAP.
Usually there is a file
/etc/openldap/slapd.conf
which contain the
configuration. The logging information will be configured
with the option loglevel
. This is a bitmap with
eleven bits today. A loglevel of 63 mean that the bits one to
five are set. A good choice is 63 for a first debugging session.
You can read the details in man slapd.conf.
Sometimes there are problems with the compilation of the Perl module
gettext. The most common problem is a loader
problem where ld reports a non existing
library intl (-lintl
).
The output look like follows:
The Perl module is only an interface to a C library from gettext. You must install the package gettext and the problem should be solved.
Sometimes you find errors releated to SET NAMES in the logs of your SQL server if you use MySQL before version 4.1. This happens because MySQL supports this part of SQL only since version 4.1. You can ignore these errormessages.
Databases and their tables are usually created for a specific language. OpenCA is able to support all available languages at all times since version 0.9.2. This requires the support for several different character encodings like UTF-8, ISO 8859-1, ISO 8859-2 and ISO 8859-15. Therefore OpenCA must set the used character encoding before it uses a database. If we don't do this then the database can abort actions because of failing transformations. SQL defines a syntax for such operation which include SET NAMES. MySQL implemented this only since 4.1 and therefore we run the SQL command in a way which don't abort transactions. Therefore you can ignore errormessages from your MySQL database if the version is smaller than 4.1.
OpenSSL's webpage. OpenSSL project.
OpenCA's webpage. OpenCA project.
PKIX. IETF.
RFC 2246 The TLS Protocol Version 1.0. IETF. January 1999.
RFC 2253 LDAP - UTF-8 String Representation of Distinguished Names. IETF. December 1997.
RFC 2311 S/MIME: Version 2 Message Specification. IETF. March 1998.
RFC 2312 S/MIME: S/MIME Version 2 Certificate Handling. IETF. March 1998.
RFC 2315 PKCS #7: Cryptographic Message Syntax Version 1.5. IETF. March 1998.
RFC 2595 Using TLS with IMAP, POP3 and ACAP. IETF. Jun 1999.
RFC 2828 Internet Security Glossary. IETF. May 2000.
RFC 2818 HTTP Over TLS. IETF. May 2000.
RFC 2898 PKCS #5: Password-Based Cryptography Specification Version 2.0. IETF. September 2000.
RFC 2985 PKCS #9: Selected Object Classes and Attribute Types Version 2.0. IETF. November 2000.
RFC 2986 PKCS #10: Certification Request Syntax Specification Version 1.7. IETF. November 2000.
RFC 3207 SMTP Service Extension for Secure SMTP over Transport Layer Security. IETF. February 2002.
RFC 3279 Algorithms and Identifiers for the Internet X.509 Public Key Infrastructure Certificate and Certificate Revocation List (CRL) Profile. IETF. April 2002.
[RFC3280] RFC 3280 Internet X.509 Public Key Infrastructure Certificate and Certificate Revocation List (CRL) Profile. IETF. April 2002.
RFC 3447 Public-Key Cryptography Standards (PKCS) #1: RSA Cryptography Specifications Version 2.1. IETF. February 2003.
OpenLDAP's webpage. OpenLDAP.
PKI - Implementing and Managing E-Security. Coriolis Group. 2001.
LDAP, Programming Directory-Enabled Applications with Lightweight Directory Access Protocol. Macmillan Technology Series, Macmillan Technical Publishing. 1997.
Understanding and Deploying LDAP Directory Services. Macmillan Network Architecture and Development Series, Macmillan Technical Publishing. 1999.
Public Key Infrastructure for KMU's (Project report, German only). Zuericher Hochschule Winterthur. Jule 2003.
Aufbau und Betrieb einer Zertifizierungsinstanz, DFN-PCA Handbuch (German only). Deutsches Forschungsnetz - DFN (Germany`s National Research and Education Network). May 2000.
Homepage of the SET standard. SET Secure Electronic Transaction LLC.
DocBook XSL: The Complete Guide. Sagehill Enterprises. September 2003.
DocBook: The Definitive Guide. O'Reilly & Associates. 31st December 2003.
The DocBook Wiki. DocBook.
OpenTSA's webpage. OpenTSA project.
OpenCA Live CD . Dartmouth PKI Lab Outreach.
Active Directory Services
Certification Authority
Certificate Revocation List
Certificate Revocation Request
Certificate Signing Request
DeMilitarized Zone is an area which is isolated from the inner and outer network of a firewall system. It is used to place servers in a protected area which has no direct access to the inner and outer network but can offer this service to people or systems in both areas. This is a very short description. Please consult specialized books or even better humans if you have absolute no idea how firewall systems work. This is really security relevant.
Distinguished Name
The Global Unique IDentifier is a 16 byte string for an object the ADS. Every domain controller have such a GUID for example and they must be present in the certificates of these domain controllers.
is nothing else than HTTP trough a SSL or TLS tunnel. It protects the communication between a http server and a browser. It was the first application for SSL and should be today the world's most widely used SSL application.
is nothing else than IMAP trough a SSL or TLS tunnel. It protects the communication between a mail server and a mail user agent if the user reads and manage it's mail.
Lightweight Directory Access Protocol
Level Of Assurance defines the quality of the identification of the certificate owner. Sometimes it is usefule to know how the owner of certificate was identified or would you send money because of signed mail if the owner was identified via email?
Mail Transfer Agent - a tool to send mail to other users or mail servers, e.g. Mozilla, Outlook (Express). Sendmail can be a MTA too if it acts as client.
Mail User Agent - e.g. Mozilla, Outlook (Express). This is the tool which a user uses to read and handle it's mail.
This is the management interface for an OpenCA installation on one machine.
Public Key Cryptography Standards are developed by RSA Security. They are widely accepted in the PKI area.
defines the ASN.1 structure of certificate signing request
is nothing else than POP trough a SSL or TLS tunnel. It protects the communication between a mail server and a mail user agent if the user reads and manage it's mail.
Registration Authority
Simple Certificate Enrollment Protocol was developed by Cisco and is used to handle the communication between a PKI and networkcomponents like router, switches and other (perhaps software) VPN components.
Secure MIME is a standard which defines how secured emails must be formatted. Please check the listed RFCs to find references to more detailed descriptions.
Simple Mail Transfer Protocol is used by mailservers like sendmail to exchange the mails. The protocol is used for mailtransfer from simple MTAs like Mozilla to servers like sendmail and for transfers from server to server.
is nothing else than SMTP trough a TLS tunnel. It protects the communication between a mail server and a mail user agent if the user reads and manage it's mail.
Signed Public Key And Challenge is a standard for CSRs from Netscape.
Secure Socket Layer is a tranport layer security protocol. It was one of the first certificate based security protocols for tunneling. Netscape developed this protocol for it's webbrowser Navigator. TLS is the standardized follower of this protocol. Today the versions 2 and 3 are still widely used and supported.
The subject of a certificate or of a request is the name of the certificate or request. The subject is a distinguished name and looks like “cn=Jon Doe, ou=Sales, o=startup, c=us”.
A symlink is nothing else than a symbolic link. Such links will be created by OpenCA usually with ln -s. We always try to avoid the shortcut but sometimes we are simply to fast ;-)
Transport Layer Security is a tranport layer security protocol. It is the standardized successor of SSL. All modern browsers use this protocol but it can be used to tunnel every other TCP based service.
Trusted Third Party is usually a trustworthy external CA.
The Universal Principle Name is the user account plus the domain name. They are connected with an at-sign @. Example: john_doe@company.com.
Table of Contents
This appendix gives an over view of the current strategic direction of the OpenCA development. As all things it is liable to change, but this is the current thinking.
A statement from the OpenCA team to say that for a given server and database environment what is the expected volume of certificates. This is important for users planning installations.
This would allow for complex scripts to be written around the OpenCA environment, hopefully paving the way for future modifications. This would also lead to the posibility of XKMS, CMS, SOAP based clients.
This would inlcude a fully documented PERL API scripting interface.
Automation of regular operations like: CRL production, certificate signing. This is important in production environments where you do not want Operations staff to have to manually produce regular CRLs, etc.
To accommodate an on-line CA model. i.e. a user can request a certificate and in the same session get the requested cert back. This can be used for "free email certs" or in closed user groups where only certain people have access to the public interface. It may be that this would only work with CA root key in hardware, or a special CA user logged on on a secure terminal to give the environment access to the CA password. In addition to this, the CA may keep a log of the number of times it was accessed.
A high risk environment mode, which is based on a cd-rom or some similar write protected media, changeable configuration data and exchange data are hold on writeable media (like usb-based-hardware, maybe encryptable), and the ossupports something like se-linux or similar.
Audit of RA and CA operations to a tamper proof signed log. This is possibly a requirement to achive any form of accreditiation.
A function that ensures OpenCA is running in a "known" environment. Perhaps md5 signature creation (after installation) and run time validation.
When reaching the end of the CA certificate lifetime, there is a certain point after which no usable end entity certificates can be issued whose desired validity *fully* fits into the CA certificates validity.
OpenCA could introduce the idea of "certificate profiles" where a user "requests" once but gets a "Profile" of certificate types". Secure storage and recovery of encryption keys would be part of this mechanism. The start of this is in the new Batch processes in the form of the "Process".
A function to provide (optional) key backup and recovery.
Enhancing the existing management screens to allow management of certificate roles and extensions, access control settings and node management i.e. a front end to the OpenSSL config files.
Screens to allow users to renew their certificates, modify DN's etc.
The ability to allow a user to request a certificate and authenticate themselves the authentication token is then checked against an independent directory.
I frequently get lost in the system when trying to debug things, often I wonder what functions get executed by OpenCA, the CGI system seems very opaque to me (and I consider myself an experienced Perl hacker)
I have seen OpenCA report crude error messages on seemingly harmless error conditions. When checking the code it was often something like an uninitialized variable that was used to call a method on.