Difference between pages "Funtoo:Keychain" and "Virtual Packages"

From Funtoo
(Difference between pages)
Jump to navigation Jump to search
 
(two first cases for virtuals)
 
Line 1: Line 1:
{{Article
Virtual packages are special packages that correspond to a feature that can be satisfied by one or more package(s). This Wiki page aims to describe when and how to use them correctly, and what are their implications.
|Subtitle=Official Project Page
|Author=Drobbins
}}
<tt>Keychain</tt> helps you to manage SSH and GPG keys in a convenient and secure manner. It acts as a frontend to <tt>ssh-agent</tt> and <tt>ssh-add</tt>, but allows you to easily have one long running <tt>ssh-agent</tt> process per system, rather than the norm of one <tt>ssh-agent</tt> per login session.


This dramatically reduces the number of times you need to enter your passphrase. With <tt>keychain</tt>, you only need to enter a passphrase once every time your local machine is rebooted. <tt>Keychain</tt> also makes it easy for remote cron jobs to securely &quot;hook in&quot; to a long running <tt>ssh-agent</tt> process, allowing your scripts to take advantage of key-based logins.
== Virtual packages, metapackages and package sets ==
Virtual packages, metapackages and package sets are similar concepts. However, they have a few important differences that make them fit for different use cases.


== Download and Resources ==
Virtual packages and metapackages are regular Funtoo packages (ebuilds) that install no files. Instead, they cause other packages to be installed by specifying them in their runtime dependencies. They can both be used in any context valid for regular packages. They can have multiple versions, slots and USE flags. They have to be located in an active repository, and once there they can be installed and uninstalled like regular packages.


The latest release of keychain is version <tt>2.7.2_beta1</tt>, and was released on July 7, 2014. The current version of keychain supports <tt>gpg-agent</tt> as well as <tt>ssh-agent</tt>.
Package sets are not packages but special atoms supported by Portage. Package sets can only specify other packages, either via a static list or dynamically (e.g. via running Python code that determines the package list). Package sets can't be versioned and don't have USE flags. Package sets can be used alongside packages in emerge commands and other package sets but they can't be referenced inside regular packages. Package sets can be installed into user's system, located in repositories or created by user in Portage configuration.


Keychain is compatible with many operating systems, including <tt>AIX</tt>, <tt>*BSD</tt>, <tt>Cygwin</tt>, <tt>MacOS X</tt>, <tt>Linux</tt>, <tt>HP/UX</tt>, <tt>Tru64 UNIX</tt>, <tt>IRIX</tt>, <tt>Solaris</tt> and <tt>GNU Hurd</tt>.
Virtual packages represent a commonly used feature that can be provided by multiple different providers. Virtuals provide a convenient way of specifying all possible alternatives without having to update multiple ebuilds.


=== Download ===
Metapackages and package sets are used to represent lists of packages that user may want to install together. They provide a convenience for users, e.g. providing a shortcut to install all packages comprising a desktop environment.


* ''Release Archive''
== When virtual packages can be used? ==
** [http://www.funtoo.org/distfiles/keychain/keychain-2.7.2_beta1.tar.bz2 keychain 2.7.2_beta1]
For virtual package ebuild to work correctly, the two following requirements must be met:
** [http://www.funtoo.org/distfiles/keychain/keychain-2.7.1.tar.bz2 keychain 2.7.1]
# the virtual providers must be interchangeable at runtime with no consequences to the reverse dependencies. In other words, installing another provider and removing the currently used provider must not cause any breakage or require reverse dependencies to be rebuilt.
# Reverse dependencies need to have consistent, predictable requirements for the alternatives. In other words, the packages must not require a very specific versions of the alternatives.


* ''Apple MacOS X Packages''
Virtuals can not be used if the underlying packages don't provide binary compatibility at least between predictable range of versions.
** [http://www.funtoo.org/distfiles/keychain/keychain-2.7.1-macosx.tar.gz keychain 2.7.1 MacOS X package]


Keychain development sources can be found in the [http://www.github.com/funtoo/keychain keychain git repository]. Please use the [https://bugs.funtoo.org Funtoo Linux bug tracker] and [irc://irc.freenode.net/funtoo #funtoo irc channel] for keychain support questions as well as bug reports.
== Common uses for virtual packages ==
=== System components and services ===
Example: ''virtual/service-manager''


=== Project History ===
One of the common uses for virtuals is to define abstract ''system services''. Those virtuals are not very specific on how those services are provided. They are mostly intended to be used in the @system package set, to ensure that the user system doesn't lack key components such as a service manager or a package manager.


Daniel Robbins originally wrote <tt>keychain</tt> 1.0 through 2.0.3. 1.0 was written around June 2001, and 2.0.3 was released in late August, 2002.
The providers for this kind of virtuals do not have to meet any specific requirements except for having a particular function. In particular, there's no requirement for common configuration or provided executables. The user is responsible for ensuring that the installed implementation is set up and working.


After 2.0.3, <tt>keychain</tt> was maintained by various Gentoo developers, including Seth Chandler, Mike Frysinger and Robin H. Johnson, through July 3, 2003.
=== Tools provided by multiple packages ===
Example: ''virtual/eject''


On April 21, 2004, Aron Griffis committed a major rewrite of <tt>keychain</tt> which was released as 2.2.0. Aron continued to actively maintain and improve <tt>keychain</tt> through October 2006 and the <tt>keychain</tt> 2.6.8 release. He also made a few commits after that date, up through mid-July, 2007. At this point, <tt>keychain</tt> had reached a point of maturity.
This kind of virtuals is used when multiple packages may provide tools with the same names. The virtual is used in packages that rely on those tools being present, in particular when the tools are used at build-time of the package or are called by package's scripts (executables).


In mid-July, 2009, Daniel Robbins migrated Aron's mercurial repository to git and set up a new project page on funtoo.org, and made a few bug fix commits to the git repo that had been collecting in [http://bugs.gentoo.org bugs.gentoo.org]. Daniel continues to maintain <tt>keychain</tt> and supporting documentation on funtoo.org, and plans to make regular maintenance releases of <tt>keychain</tt> as needed.
While the tools don't necessarily need to be fully compatible, they need to have a common basic usage. In particular, when a tool from one provider is replaced by a tool from another, the reverse dependencies must remain in working state, with no need for rebuilds or configuration adjustments.
 
== Quick Setup ==
 
=== Linux ===
 
To install under Gentoo or Funtoo Linux, type
<console>
###i## emerge keychain
</console>
 
For other Linux distributions, use your distribution's package manager, or download and install using the source tarball above. Then generate RSA/DSA keys if necessary. The quick install docs assume you have a DSA key pair named <tt>id_dsa</tt> and <tt>id_dsa.pub</tt> in your <tt>~/.ssh/</tt> directory. Add the following to your <tt>~/.bash_profile</tt>:
 
{{file|name=~/.bash_profile|body=
eval `keychain --eval --agents ssh id_rsa`
}}
 
If you want to take advantage of GPG functionality, ensure that GNU Privacy Guard is installed and omit the <tt>--agents ssh</tt> option above.
 
=== Apple MacOS X ===
 
To install under MacOS X, install the MacOS X package for keychain. Assuming you have an <tt>id_dsa</tt> and <tt>id_dsa.pub</tt> key pair in your <tt>~/.ssh/</tt> directory, add the following to your <tt>~/.bash_profile</tt>:
 
{{file|name=~/.bash_profile|body=
eval `keychain --eval --agents ssh --inherit any id_dsa`
}}
 
{{Fancynote|The <tt>--inherit any</tt> option above causes keychain to inherit any ssh key passphrases stored in your Apple MacOS Keychain. If you would prefer for this to not happen, then this option can be omitted.}}
 
== Background ==
 
You're probably familiar with <tt>ssh</tt>, which has become a secure replacement for the venerable <tt>telnet</tt> and <tt>rsh</tt> commands.
 
Typically, when one uses <tt>ssh</tt> to connect to a remote system, one supplies a secret passphrase to <tt>ssh</tt>, which is then passed in encrypted form over the network to the remote server. This passphrase is used by the remote <tt>sshd</tt> server to determine if you should be granted access to the system.
 
However, OpenSSH and nearly all other SSH clients and servers have the ability to perform another type of authentication, called asymmetric public key authentication, using the RSA or DSA authentication algorithms. They are very useful, but can also be complicated to use. <tt>keychain</tt> has been designed to make it easy to take advantage of the benefits of RSA and DSA authentication.
 
== Generating a Key Pair ==
 
To use RSA and DSA authentication, first you use a program called <tt>ssh-keygen</tt> (included with OpenSSH) to generate a ''key pair'' -- two small files. One of the files is the ''public key''. The other small file contains the ''private key''. <tt>ssh-keygen</tt> will ask you for a passphrase, and this passphrase will be used to encrypt your private key. You will need to supply this passphrase to use your private key. If you wanted to generate a DSA key pair, you would do this:
 
<console># ##i##ssh-keygen -t dsa
Generating public/private dsa key pair.</console>
You would then be prompted for a location to store your key pair. If you do not have one currently stored in <tt>~/.ssh</tt>, it is fine to accept the default location:
 
<console>Enter file in which to save the key (/root/.ssh/id_dsa): </console>
Then, you are prompted for a passphrase. This passphrase is used to encrypt the ''private key'' on disk, so even if it is stolen, it will be difficult for someone else to use it to successfully authenticate as you with any accounts that have been configured to recognize your public key.
 
Note that conversely, if you '''do not''' provide a passphrase for your private key file, then your private key file '''will not''' be encrypted. This means that if someone steals your private key file, ''they will have the full ability to authenticate with any remote accounts that are set up with your public key.''
 
Below, I have supplied a passphrase so that my private key file will be encrypted on disk:
 
<console>Enter passphrase (empty for no passphrase): ##i#########
Enter same passphrase again: ##i#########
Your identification has been saved in /var/tmp/id_dsa.
Your public key has been saved in /var/tmp/id_dsa.pub.
The key fingerprint is:
5c:13:ff:46:7d:b3:bf:0e:37:1e:5e:8c:7b:a3:88:f4 root@devbox-ve
The key's randomart image is:
+--[ DSA 1024]----+
|          .      |
|          o  . |
|          o . ..o|
|      . . . o  +|
|        S    o. |
|            . o.|
|        .  ..++|
|        . o . =o*|
|        . E .+*.|
+-----------------+</console>
 
== Setting up Authentication ==
 
Here's how you use these files to authenticate with a remote server. On the remote server, you would append the contents of your ''public key'' to the <tt>~.ssh/authorized_keys</tt> file, if such a file exists. If it doesn't exist, you can simply create a new <tt>authorized_keys</tt> file in the remote account's <tt>~/.ssh</tt> directory that contains the contents of your local <tt>id_dsa.pub</tt> file.
 
Then, if you weren't going to use <tt>keychain</tt>, you'd perform the following steps. On your local client, you would start a program called <tt>ssh-agent</tt>, which runs in the background. Then you would use a program called <tt>ssh-add</tt> to tell <tt>ssh-agent</tt> about your secret private key. Then, if you've set up your environment properly, the next time you run <tt>ssh</tt>, it will find <tt>ssh-agent</tt> running, grab the private key that you added to <tt>ssh-agent</tt> using <tt>ssh-add</tt>, and use this key to authenticate with the remote server.
 
Again, the steps in the previous paragraph is what you'd do if <tt>keychain</tt> wasn't around to help. If you are using <tt>keychain</tt>, and I hope you are, you would simply add the following line to your <tt>~/.bash_profile</tt> or if a regular user to<tt>~/.bashrc</tt> :
 
{{file|name=~/.bash_profile|body=
eval `keychain --eval id_dsa`
}}
 
The next time you log in or source your <tt>~/.bash_profile</tt> or if you use <tt>~/.bashrc</tt>, <tt>keychain</tt> will start, start <tt>ssh-agent</tt> for you if it has not yet been started, use <tt>ssh-add</tt> to add your <tt>id_dsa</tt> private key file to <tt>ssh-agent</tt>, and set up your shell environment so that <tt>ssh</tt> will be able to find <tt>ssh-agent</tt>. If <tt>ssh-agent</tt> is already running, <tt>keychain</tt> will ensure that your <tt>id_dsa</tt> private key has been added to <tt>ssh-agent</tt> and then set up your environment so that <tt>ssh</tt> can find the already-running <tt>ssh-agent</tt>. It will look something like this:
 
Note that when <tt>keychain</tt> runs for the first time after your local system has booted, you will be prompted for a passphrase for your private key file if it is encrypted. But here's the nice thing about using <tt>keychain</tt> -- even if you are using an encrypted private key file, you will only need to enter your passphrase when your system first boots (or in the case of a server, when you first log in.) After that, <tt>ssh-agent</tt> is already running and has your decrypted private key cached in memory. So if you open a new shell, you will see something like this:
 
This means that you can now <tt>ssh</tt> to your heart's content, without supplying a passphrase.
 
You can also execute batch <tt>cron</tt> jobs and scripts that need to use <tt>ssh</tt> or <tt>scp</tt>, and they can take advantage of passwordless RSA/DSA authentication as well. To do this, you would add the following line to the top of a bash script:
 
{{file|name=example-script.sh|body=
eval `keychain --noask --eval id_dsa` || exit 1
}}
 
The extra <tt>--noask</tt> option tells <tt>keychain</tt> that it should not prompt for a passphrase if one is needed. Since it is not running interactively, it is better for the script to fail if the decrypted private key isn't cached in memory via <tt>ssh-agent</tt>.
 
== Keychain Options ==
 
=== Specifying Agents ===
 
In the images above, you will note that <tt>keychain</tt> starts <tt>ssh-agent</tt>, but also starts <tt>gpg-agent</tt>. Modern versions of <tt>keychain</tt> also support caching decrypted GPG keys via use of <tt>gpg-agent</tt>, and will start <tt>gpg-agent</tt> by default if it is available on your system. To avoid this behavior and only start <tt>ssh-agent</tt>, modify your <tt>~/.bash_profile</tt> as follows:
 
{{file|name=~/.bash_profile|body=
eval `keychain --agents ssh --eval id_dsa` || exit 1
}}
 
The additional <tt>--agents ssh</tt> option tells <tt>keychain</tt> just to manage <tt>ssh-agent</tt>, and ignore <tt>gpg-agent</tt> even if it is available.
 
=== Clearing Keys ===
 
Sometimes, it might be necessary to flush all cached keys in memory. To do this, type:
 
<console># ##i##keychain --clear</console>
Any agent(s) will continue to run.
 
=== Improving Security ===
 
To improve the security of <tt>keychain</tt>, some people add the <tt>--clear</tt> option to their <tt>~/.bash_profile</tt> <tt>keychain</tt> invocation. The rationale behind this is that any user logging in should be assumed to be an intruder until proven otherwise. This means that you will need to re-enter any passphrases when you log in, but cron jobs will still be able to run when you log out.
 
=== Stopping Agents ===
 
If you want to stop all agents, which will also of course cause your keys/identities to be flushed from memory, you can do this as follows:
 
<console># ##i##keychain -k all</console>
If you have other agents running under your user account, you can also tell <tt>keychain</tt> to just stop only the agents that <tt>keychain</tt> started:
 
<console># ##i##keychain -k mine</console>
 
=== GPG ===
 
Keychain can ask you for your GPG passphrase if you provide it the GPG key ID. To find it out:
<console>
$##i## gpg -k
pub  2048R/DEADBEEF 2012-08-16
uid                  Name (Comment) <email@host.tld>
sub  2048R/86D2FAC6 2012-08-16
</console>
 
Note the '''DEADBEEF''' above is the ID. Then, in your login script, do your usual
 
<console>
$##i## keychain --dir ~/.ssh/.keychain ~/.ssh/id_rsa DEADBEEF
$##i## source ~/.ssh/.keychain/$HOST-sh
$##i## source ~/.ssh/.keychain/$HOST-sh-gpg
</console>
 
=== Learning More ===
 
The instructions above will work on any system that uses <tt>bash</tt> as its default shell, such as most Linux systems and Mac OS X.
 
To learn more about the many things that <tt>keychain</tt> can do, including alternate shell support, consult the keychain man page, or type <tt>keychain --help | less</tt> for a full list of command options.
 
I also recommend you read my original series of articles about [http://www.openssh.com OpenSSH] that I wrote for IBM developerWorks, called <tt>OpenSSH Key Management</tt>. Please note that <tt>keychain</tt> 1.0 was released along with Part 2 of this article, which was written in 2001. <tt>keychain</tt> has changed quite a bit since then. In other words, read these articles for the conceptual and [http://www.openssh.com OpenSSH] information, but consult the <tt>keychain</tt> man page for command-line options and usage instructions :)
 
* [http://www.ibm.com/developerworks/library/l-keyc.html Common Threads: OpenSSH key management, Part 1] - Understanding RSA/DSA Authentication
* [http://www.ibm.com/developerworks/library/l-keyc2/ Common Threads: OpenSSH key management, Part 2] - Introducing <tt>ssh-agent</tt> and <tt>keychain</tt>
* [http://www.ibm.com/developerworks/library/l-keyc3/ Common Threads: OpenSSH key management, Part 3] - Agent forwarding and <tt>keychain</tt> improvements
 
As mentioned at the top of the page, <tt>keychain</tt> development sources can be found in the [http://www.github.com/funtoo/keychain keychain git repository]. Please use the [http://groups.google.com/group/funtoo-dev funtoo-dev mailing list] and [irc://irc.freenode.net/funtoo #funtoo irc channel] for keychain support questions as well as bug reports.
 
[[Category:HOWTO]]
[[Category:Projects]]
[[Category:First Steps]]
[[Category:Articles]]
{{ArticleFooter}}

Revision as of 13:31, February 7, 2015

Virtual packages are special packages that correspond to a feature that can be satisfied by one or more package(s). This Wiki page aims to describe when and how to use them correctly, and what are their implications.

Virtual packages, metapackages and package sets

Virtual packages, metapackages and package sets are similar concepts. However, they have a few important differences that make them fit for different use cases.

Virtual packages and metapackages are regular Funtoo packages (ebuilds) that install no files. Instead, they cause other packages to be installed by specifying them in their runtime dependencies. They can both be used in any context valid for regular packages. They can have multiple versions, slots and USE flags. They have to be located in an active repository, and once there they can be installed and uninstalled like regular packages.

Package sets are not packages but special atoms supported by Portage. Package sets can only specify other packages, either via a static list or dynamically (e.g. via running Python code that determines the package list). Package sets can't be versioned and don't have USE flags. Package sets can be used alongside packages in emerge commands and other package sets but they can't be referenced inside regular packages. Package sets can be installed into user's system, located in repositories or created by user in Portage configuration.

Virtual packages represent a commonly used feature that can be provided by multiple different providers. Virtuals provide a convenient way of specifying all possible alternatives without having to update multiple ebuilds.

Metapackages and package sets are used to represent lists of packages that user may want to install together. They provide a convenience for users, e.g. providing a shortcut to install all packages comprising a desktop environment.

When virtual packages can be used?

For virtual package ebuild to work correctly, the two following requirements must be met:

  1. the virtual providers must be interchangeable at runtime with no consequences to the reverse dependencies. In other words, installing another provider and removing the currently used provider must not cause any breakage or require reverse dependencies to be rebuilt.
  2. Reverse dependencies need to have consistent, predictable requirements for the alternatives. In other words, the packages must not require a very specific versions of the alternatives.

Virtuals can not be used if the underlying packages don't provide binary compatibility at least between predictable range of versions.

Common uses for virtual packages

System components and services

Example: virtual/service-manager

One of the common uses for virtuals is to define abstract system services. Those virtuals are not very specific on how those services are provided. They are mostly intended to be used in the @system package set, to ensure that the user system doesn't lack key components such as a service manager or a package manager.

The providers for this kind of virtuals do not have to meet any specific requirements except for having a particular function. In particular, there's no requirement for common configuration or provided executables. The user is responsible for ensuring that the installed implementation is set up and working.

Tools provided by multiple packages

Example: virtual/eject

This kind of virtuals is used when multiple packages may provide tools with the same names. The virtual is used in packages that rely on those tools being present, in particular when the tools are used at build-time of the package or are called by package's scripts (executables).

While the tools don't necessarily need to be fully compatible, they need to have a common basic usage. In particular, when a tool from one provider is replaced by a tool from another, the reverse dependencies must remain in working state, with no need for rebuilds or configuration adjustments.