Merge 'mirroring-revamp' into pu

This commit is contained in:
Sitaram Chamarty 2011-08-29 13:55:21 +05:30
commit 2eee4d627a
24 changed files with 1071 additions and 459 deletions

View file

@ -45,7 +45,6 @@ $GL_NO_DAEMON_NO_GITWEB = 0;
# $GL_ALL_READ_ALL = 0;
$GIT_PATH="";
$GL_GITCONFIG_KEYS = "";
# $GL_GITCONFIG_WILD = 0;
$GL_NO_CREATE_REPOS = 0;
$GL_NO_SETUP_AUTHKEYS = 0;
# $GL_WILDREPOS_DEFPERMS = 'R @all';
@ -67,6 +66,7 @@ $SVNSERVE = "";
# PLEASE USE SINGLE QUOTES ABOVE, NOT DOUBLE QUOTES
$GL_WILDREPOS_PERM_CATS = "READERS WRITERS";
# $GL_SITE_INFO = "XYZ.COM DEVELOPERS: PLEASE SEE http://xyz.com/gitolite/help first";
# $GL_HOSTNAME = "frodo"; # read doc/mirroring.mkd COMPLETELY before setting this
# ------------------------------------------------------------------------------
# rarely changed variables

View file

@ -444,15 +444,21 @@ Gitolite allows you to specify access for git-daemon and gitweb. See
(Thanks to teemu dot matilainen at iki dot fi)
Sometimes you want to specify `git config` settings for some of your repos.
For example, you may have a custom post-receive hook that sends an email when
a push happens, and this hook needs to know whom to send the email to, etc.
> ----
You can set git config values by specifying something like this within a
"repo" paragraph:
> **Note**: this won't work unless the rc file has the right settings;
> please see `$GL_GITCONFIG_KEYS` in [doc/gitolite.rc.mkd][rcdoc] for
> details and security information.
example usage: if you placed a hook in hooks/common that requires
configuration information that is specific to each repo, you could do this:
> ----
Sometimes you want to specify `git config` settings for your repos.
For example, say you have a custom post-receive hook that sends an email when
a push happens, and this hook looks in the config for whom to send the email
to, etc.
You can set these git config values within a "repo" paragraph:
repo gitolite
config hooks.mailinglist = gitolite-commits@example.tld
@ -466,11 +472,39 @@ The syntax is simple:
This does either a plain "git config section.key value" (for the first 3
examples above) or "git config --unset-all section.key" (for the last
example). Other forms (--add, the `value_regex`, etc) are not supported.
example). Other forms of the `git config` command (`--add`, the
`value_regex`, etc) are not supported.
**Note**: this won't work unless the rc file has the right settings; please
see comments around the variable `$GL_GITCONFIG_KEYS` in doc/gitolite.rc.mkd
for details and security information.
> ----
> **WARNING**: simply deleting the config line from the `conf/gitolite.conf`
> file will *not* delete the variable from `repo.git/config`. The syntax in
> the last example is the *only* way to make gitolite execute a
> `--unset-all` operation on the given key.
> ----
You can repeat the 'config' line as many times as you like, and the last
occurrence will be the one in effect. This allows you to override settings
just for one project, as in this example:
repo @all
config gitolite.mirror.master = "frodo"
config gitolite.mirror.slaves = "sam gollum"
repo top-secret-project
# only sam, because we don't trust gollum
config gitolite.mirror.slaves = "sam"
The "delete config variable" syntax can also be used, if you wish:
repo highlander # there can be only one!
config gitolite.mirror.master =
config gitolite.mirror.slaves =
As you can see, the general idea is to place the most generic ones (`repo
@all`, or repo patterns like `repo foo.*`) first, and place more specific ones
later to override the generic settings.
[rcdoc]: http://sitaramc.github.com/gitolite/doc/gitolite.rc.html

View file

@ -154,12 +154,6 @@ on feedback from my users to find or fix issues.
`\\.` (two backslashes and a dot). So this is how you'd allow any keys in
the "foo" category: `$GL_GITCONFIG_KEYS = "foo\\..*";`
* `$GL_GITCONFIG_WILD`, boolean, default 0
This setting allows gitconfig keys even for wild repos. This is an
efficiency issue more than a security issue, since this requires trawling
through all of `$REPO_BASE` looking for stuff :)
* `$GL_NO_CREATE_REPOS`, boolean, default 0
DO NOT CHANGE THIS unless you have other means to create repos and

View file

@ -1,52 +1,65 @@
## mirroring a gitolite setup
# mirroring gitolite servers
Mirroring git repos is essentially a one-liner. For each mirror you want to
update, you just add a post-receive hook that says
Mirroring a repo is simple in git; you just need code like this in a
`post-receive` hook in each repo:
#!/bin/bash
git push --mirror slave_user@mirror.host:/path/to/repo.git
# if running gitolite, the $GL_REPO variable could be useful:
# git push --mirror slave_user@mirror.host:/repo/base/path/$GL_REPO.git
But life is never that simple...
**This document has been tested using a 3-server setup, all installed using
the "non-root" method (see doc/1-INSTALL.mkd). However, the process is
probably not going to be very forgiving of human error -- like anything that
is this deep in "system admin" territory, errors are likely to be costly. If
you're the kind who hits enter first and then thinks about what he typed,
you're in for some fun times ;-)**
**On the plus side, everything we do is done using git commands, so things are
never *really* lost until you do a `git gc`**.
----
**Update 2011-03-10**: I wrote this with a typical "corporate" setup in mind
where all the servers involved are owned and administered by the same group of
people. As a result, the scripts assume the servers trust each other
completely. If that is not your situation, you will have to add code into
`gl-mirror-shell` to limit the commands the remote may send. Patches welcome
:-)
For a lot of people, though, mirroring is more than just 'backup', and their
needs are complex enough that setup is hard.
----
In this document:
* <a href="#_why">why</a>
* <a href="#_RULE_NUMBER_ONE_">RULE NUMBER ONE!</a>
* <a href="#_things_that_will_NOT_be_mirrored_by_this_process">things that will NOT be mirrored by this process</a>
* <a href="#_conventions_in_this_document">conventions in this document</a>
* <a href="#_setting_up_mirroring">setting up mirroring</a>
* <a href="#_install_gitolite_on_all_servers">install gitolite on all servers</a>
* <a href="#_generate_keypairs">generate keypairs</a>
* <a href="#_setup_the_mirror_shell_on_each_server">setup the mirror-shell on each server</a>
* <a href="#_set_slaves_to_slave_mode">set slaves to slave mode</a>
* <a href="#_set_slave_server_lists">set slave server lists</a>
* <a href="#_efficiency_versus_paranoia">efficiency versus paranoia</a>
* <a href="#_syncing_the_mirrors_the_first_time">syncing the mirrors the first time</a>
* <a href="#_switching_over">switching over</a>
* <a href="#_the_return_of_foo">the return of foo</a>
* <a href="#_switching_back">switching back</a>
* <a href="#_making_foo_a_slave">making foo a slave</a>
* <a href="#_URLs_that_your_users_will_use">URLs that your users will use</a>
* <a href="#_IMPORTANT_cautions">IMPORTANT cautions</a>
* <a href="#_concepts_and_terminology">concepts and terminology</a>
* <a href="#_setup_and_usage">setup and usage</a>
* <a href="#_server_level_setup">server level setup</a>
* <a href="#_repository_level_setup">repository level setup</a>
* <a href="#_commands_to_re_sync_mirrors">commands to (re-)sync mirrors</a>
* <a href="#_details">details</a>
* <a href="#_the_conf_gitolite_conf_file">the `conf/gitolite.conf` file</a>
* <a href="#_redirecting_pushes">redirecting pushes</a>
* <a href="#_discussion">discussion</a>
* <a href="#_problems_with_the_old_mirroring_model">problems with the old mirroring model</a>
* <a href="#_the_new_mirroring_model">the new mirroring model</a>
* <a href="#_appendix_A_example_cronjob_based_mirroring">appendix A: example cronjob based mirroring</a>
* <a href="#_appendix_B_efficiency_versus_paranoia">appendix B: efficiency versus paranoia</a>
----
<a name="_why"></a>
### why
This document is useful if:
* you have multiple repositories spread across multiple sites around the
country/world, and would like developers to access their local server
instead of hitting the WAN, at least for 'fetch' operations.
* you don't want all your repos mirrored to all the servers for various
reasons, technical or otherwise (epecially true when some of the mirrors
don't belong to you).
* you want some mirrors to be updated only at certain times of the day,
(with a simple command), instead of every time a push happens.
* you don't want *one* server being the master server for *all* repos;
instead you want to choose where a repo gets "mastered" based on where the
majority of that repo's users are.
* you might even, if your servers are all in your control, want the
convenience of them *pushing to a mirror*, and having the push redirect
transparently to the master server.
As you can see, this is a bit more than a backup solution ;-)
<a name="_RULE_NUMBER_ONE_"></a>
@ -62,285 +75,586 @@ Corollary: if the primary went down and you effected a changeover, you must
make sure that the primary does not come up in a push-enabled mode when it
recovers.
<a name="_things_that_will_NOT_be_mirrored_by_this_process"></a>
<a name="_IMPORTANT_cautions"></a>
### things that will NOT be mirrored by this process
### IMPORTANT cautions
Let's get this out of the way. This procedure will only mirror your git
repositories, using `git push --mirror`. Therefore, certain files will not be
mirrored:
* For reasons given in the 'discussion' section later, the mirroring process
will never *create* a repo on the receiving side. It has to exist, and be
willing to accept pushes from the master.
* gitolite log files
* "gl-creator" and "gl-perms" files
* "projects.list", "description", and entries in the "config" files within
each repo
In particular, this means that repositories created by end-users ("wild"
repos) *need to be explicitly created* on the mirror (preferably by the
same user, assuming his ssh key works there as well). Once the repo has
been created on the slave, subsequent pushes will be mirrored correctly.
None of these affect actual repo contents of course, but they could be
important, (especially the gl-creator, although if your wildcard pattern had
"CREATOR" in it you can recreate those files easily enough anyway).
* This process will *only* mirror your git repositories, using `git push
--mirror`. It will *not* mirror log files, and repo-specific files like
`gl-creater` and `gl-perms` files, or indeed anything that was manually
created or added (for example, custom config entries added manually
instead of via gitolite).
Your best bet is to use rsync for the log files, and tar for the others, at
regular intervals.
None of these affect actual repo contents of course, but they could be
important, (especially the gl-creator, although if your wildcard pattern
had "CREATOR" in it you can recreate those files easily enough anyway).
<a name="_conventions_in_this_document"></a>
* This document has been tested using a 3-server setup, all installed using
the *non-root* method (see doc/1-INSTALL.mkd). However, the process is
probably not going to be very forgiving of human error -- like anything
that is this deep in "system admin" territory, errors are likely to be
costly. If you're the kind who hits enter first and then thinks about
what he typed, you're in for some fun times ;-)
### conventions in this document
On the plus side, everything we do is done using git commands, so things
are never *really* lost until you do a `git gc`.
The userid hosting gitolite is `gitolite` on all machines. The servers are
foo, bar, and baz. At the beginning, foo is the master, the other 2 are
slaves.
* Mirroring has *not* been, and will not be, tested with gitolite installed
using the deprecated 'from-client' method. Please use one of the other
methods.
<a name="_setting_up_mirroring"></a>
* Also, this has *not* been tested with smart-http. I'm not even sure it'll
work; http is very fiddly to get right. If you want mirroring, at least
your server-to-server comms should be over ssh.
### setting up mirroring
* Finally, this method uses repo-specific `git config` variables to store
the mirroring information. Please read the **WARNING** in the
documentation on [git config commands][rsgc] if you wish to **delete** one
of those lines.
<a name="_install_gitolite_on_all_servers"></a>
[rsgc]: http://sitaramc.github.com/gitolite/doc/gitolite.conf.html#_repo_specific_git_config_commands
#### install gitolite on all servers
<a name="_concepts_and_terminology"></a>
* before running the final step in the install sequence, make sure you go to
the `hooks/common` directory and rename `post-receive.mirrorpush` to
`post-receive`. See doc/hook-propagation.mkd if you're not sure where you
should look for `hooks/common`.
### concepts and terminology
* if the server already has gitolite installed, use the normal methods to
make sure this hook gets in.
Servers can host 3 kinds of repos: master, slave, and local.
* Use the same "admin key" on all the machines, so that the same person has
gitolite-admin access to all of them.
* A repo can be a **master** on one and only one server. A repo on its
"master" server is a **native** repo, on slaves it is "non-native".
<a name="_generate_keypairs"></a>
* A **slave** repo cannot be pushed to by a user. It will only accept
pushes from a master server. (But see later for an exception).
#### generate keypairs
* A **local** repo is not involved in mirroring at all, in either direction.
Each server will be potentially logging on to one or more of the other
servers, so first generate keypairs on each of them (`ssh-keygen`) and copy
the `.pub` files to all other servers, named appropriately. So foo will have
bar.pub and baz.pub, etc.
<a name="_setup_and_usage"></a>
<a name="_setup_the_mirror_shell_on_each_server"></a>
### setup and usage
#### setup the mirror-shell on each server
<a name="_server_level_setup"></a>
XXX review this document after testing mirroring...
#### server level setup
If you installed gitolite using the from client method, run the following:
To start with, assign each server a short name. We will use 'frodo', 'sam',
and 'gollum' as examples here.
# on foo
export GL_BINDIR=$HOME/.gitolite/src
cat bar.pub baz.pub |
sed -e 's,^,command="'$GL_BINDIR'/gl-mirror-shell" ,' >> ~/.ssh/authorized_keys
1. Generate ssh keys on each machine. Copy the `.pub` files to all other
machines with the appropriate names. I.e., frodo should have sam.pub and
gollum.pub, etc.
If you installed using any of the other 3 methods do this:
2. Install gitolite on all servers, under some 'hosting user' (we'll use
`git` in our examples here). You need not use the same hosting user on
all machines.
# on foo
export GL_BINDIR=`gl-query-rc GL_BINDIR`
cat bar.pub baz.pub |
sed -e 's,^,command="'$GL_BINDIR'/gl-mirror-shell" ,' >> ~/.ssh/authorized_keys
It is not necessary to use the same "admin key" on all the machines.
However, if you do plan to mirror the gitolite-admin repo also, they will
eventually become the same anyway. In our example, frodo does mirror the
admin repo to sam, but not to gollum. (Can you really see frodo or sam
trusting gollum?)
Also do the same thing on the other machines.
3. Now copy `hooks/common/post-receive.mirrorpush` from the gitolite source,
and install it as a custom hook called `post-receive`; see [here][ch] for
instructions.
Now test this access:
4. Edit `~/.gitolite.rc` on each machine and add/edit the following lines.
The `GL_HOSTNAME` variable **must** have the correct name for that host
(frodo, sam, or gollum), so that will definitely be different on each
server. The other line can be the same, or may have additional patterns
for other `git config` keys you have previously enabled. See [here][rsgc]
and the description for `GL_GITCONFIG_KEYS` in [this][vsi] for details.
# on foo
ssh gitolite@bar pwd
# should print /home/gitolite/repositories
ssh gitolite@bar uname -a
# should print the appropriate info for that server
$GL_HOSTNAME = 'frodo'; # will be different on each server!
$GL_GITCONFIG_KEYS = "gitolite.mirror.*";
Similarly test the other combinations.
(Remember the "rc" file is NOT mirrored; it is meant to be site-local).
<a name="_set_slaves_to_slave_mode"></a>
Note: if `GL_HOSTNAME` is undefined, you cannot push to repos which have
the 'gitolite.mirror.master' config variable set. (See 'details' section
below for more info on this variable).
#### set slaves to slave mode
5. On each machine, add the keys for all other machines. For example, on
frodo you'd run these two commands:
Set slave mode on all the *slave* servers by setting `$GL_SLAVE_MODE = 1`
(uncommenting the line if necessary).
gl-tool add-mirroring-peer sam.pub
gl-tool add-mirroring-peer gollum.pub
Leave the master server's file as is.
6. Create "host" aliases on each machine to refer to all other machines. See
[here][ha] for what/why/how.
<a name="_set_slave_server_lists"></a>
The host alias for a host (in other machines' `~/.ssh/config` files) MUST
be the same as the `GL_HOSTNAME` in the referred host's `~/.gitolite.rc`.
Gitolite mirroring **requires** this consistency in naming; things will
NOT work otherwise.
#### set slave server lists
For example, if machine A's `~/.gitolite.rc` says `$GL_HOSTNAME =
'frodo';`, then all other machines must use a host alias of "frodo" in
their `~/.ssh/config` files to refer to machine A.
On the master (foo), set the names of the slaves by editing the
`~/.gitolite.rc` to contain:
Once you've done this, each host should be able to reach the other hosts and
get a response back. For example, running this on sam:
$ENV{GL_SLAVES} = 'gitolite@bar gitolite@baz';
ssh frodo info
**Note the syntax well; this is critical**:
should get you
* **this must be in single quotes** (or you must remember to escape the `@`)
* the variable is an ENV var, not a plain perl var
* the values are *space separated*
* each value represents the userid and hostname for one server
Hello sam, I am frodo.
The basic idea is that this string, should be usable in both the following
syntaxes:
Check this command from *everywhere to everywhere else*, and make sure you get
expected results. **Do NOT proceed otherwise.**
git clone gitolite@bar:repo
ssh gitolite@bar pwd
<a name="_repository_level_setup"></a>
You can also use ssh host aliases. Let's say server "bar" has a non-standard
port number:
#### repository level setup
# in ~/.ssh/config on foo
host mybar
hostname bar
user gitolite
port 2222
Setting up mirroring at the repository level instead of at the "entire server"
level gives you a lot of flexibility (see "discussion" section below).
# in ~/.gitolite.rc on foo
$ENV{GL_SLAVES} = 'bar gitolite@baz';
The basic idea is to use `git config` variables within each repo (gitolite
allows you to create them from within the gitolite.conf file so that's
convenient), and use these to specify which machine is the master and which
machines are slaves for the repo.
And that's really all there is, unless...
<font color="gray">
<a name="_efficiency_versus_paranoia"></a>
> Side note: if you just want to **simulate the old mirroring scheme**,
> despite its limitations, it's very easy. Say frodo is the master for all
> repos, and the other 2 are slaves. Just clone the gitolite-admin repos of
> all servers, add these lines to the top of each:
### efficiency versus paranoia
repo @all
config gitolite.mirror.master = "frodo"
config gitolite.mirror.slaves = "sam gollum"
> then commit, and push all 3. Finally, make a dummy commit on just the
> frodo clone and push again. At this point you can do a one-time manual
> sync (see Appendix A) if you wish but otherwise you're done.
</font>
Let's say frodo and sam are internal servers, while gollum is an external (and
therefore less trusted) server that has agreed to help us out by mirroring one
of our high traffic repos. We want the following setup:
* the "gitolite-admin" repo, as well as an internal project repo called
"ip1", should be mastered on frodo and mirrored to sam.
* internal project "ip2" has almost all of its developers closer to sam, so
it should be mastered there, and mirrored on frodo.
* an open source project we manage, "os1", should be mastered on frodo and
mirrored on both sam and gollum.
So here's how our example would go:
1. Clone frodo's and sam's gitolite-admin repos to your workstation, then add
the following lines to both their gitolite.conf files:
repo ip1 gitolite-admin
config gitolite.mirror.master = "frodo"
config gitolite.mirror.slaves = "sam"
repo ip2
config gitolite.mirror.master = "sam"
config gitolite.mirror.slaves = "frodo"
You also need normal access control lines for ip1 and ip2; I'm assuming
you already have them elsewhere, at least on frodo. (What you have on sam
won't matter in a few minutes, as you will see!)
Commit and push these changes.
2. There are a couple of quirks to keep in mind when you make changes to the
gitolite-admin repo's config.
* the first push will create the `git config` entries required, but by
then it is too late to *act* on them; i.e., actually do the mirroring.
If there were any older values, like a different list of slaves
perhaps, then those would be in effect.
This is largely because git invokes post-receive before post-update.
In theory I can work around this but I do not intend to.
Anyway, this means that after the 2 pushes, you have to make a dummy
push from frodo:
git commit --allow-empty -m empty; git push
which gets you something like this amidst the other messages:
remote: (25158&) frodo ==== (gitolite-admin) ===> sam
telling you that frodo is sending gitolite-admin to sam in the
background.
* the second quirk is that your clone of server sam's gitolite-admin
repo is now completely out of date, since frodo has overwritten it on
the server. You have to 'cd' to that clone and do this:
git fetch
git reset --hard origin/master
2. That completes the setup of the gitolite-admin and the internal project
repos. We'll now setup things for the open source project, "os1".
On frodo's gitolite-admin clone, add the following lines to
`conf/gitolite.conf`, then commit and push:
repo os1
config gitolite.mirror.master = "frodo"
config gitolite.mirror.slaves = "sam gollum"
Also, send the same lines to gollum's administrator and ask him to add
them into his conf/gitolite.conf file, commit, and push.
<a name="_commands_to_re_sync_mirrors"></a>
#### commands to (re-)sync mirrors
You don't have to put all the slaves in `gitolite.mirror.slaves`. For
example, let's say you have some repos that are very active, and two of your
mirrors that are halfway across the world are getting pushed very frequently.
But you don't need those mirrors to be that closely updated, perhaps *because*
they are halfway across the world and those guys are asleep ;-)
Or maybe there was a network glitch and even the default slaves are now
lagging, so they need to be manually synced.
Or a slave realised that one of its repos is lagging for some reason, and
wants to request an immediate update.
Whatever the reason, you need ways to sync a repo from a command line. Here
are ways to do that:
1. On the master server, you can start a **background** job to mirror a repo.
The command/syntax is
gl-mirror-shell request-push reponame [list of keys/slaves]
The list at the end is optional, and can be a mix of slave names or your
own gitolite mirror config keys. (Yes, you can have any key, named
anything you like, as long as it starts with `gitolite.mirror.`).
If the list is not supplied, the `gitolite.mirror.slaves` key is used.
Keys can have values that in turn contain a list of keys/slaves. The list
is recursively *expanded* but recursion is not *detected*. Order is
preserved while duplicates are removed. If you didn't get that, see the
example :-)
**Warning**: the `gitolite.mirror.slaves` key should have only hosts, no
keys, in it.
The program exits with a return value of "1" if it found no slaves in the
list passed, otherwise it fires off the background job, prints an
informative message, and exits with a return value of "0".
We'll take an example. Let's say your gitolite config file has this:
repo ip1
config gitolite.mirror.master = "frodo"
config gitolite.mirror.slaves = "sam merry pippin"
config gitolite.mirror.hourly = "sam legolas"
config gitolite.mirror.nightly = "gitolite.mirror.hourly gimli"
config gitolite.mirror.all = "gitolite.mirror.nightly gitolite.mirror.hourly gitolite.mirror.slaves"
Then the following commands have the results described in comments:
gl-mirror-shell request-push ip1
# which is the same as:
gl-mirror-shell request-push ip1 gitolite.mirror.slaves
# pushes to sam, merry, pippin
gl-mirror-shell request-push ip1 gollum
# pushes only to gollum. Note that gollum is not a member of any of
# the slave lists we defined.
gl-mirror-shell request-push ip1 gitolite.mirror.slaves gollum
# pushes to sam, merry, pippin, gollum
gl-mirror-shell request-push ip1 gitolite.mirror.slaves gitolite.mirror.hourly
# pushes to sam, merry, pippin, legolas
gl-mirror-shell request-push ip1 gitolite.mirror.all
# pushes to sam, legolas, gimli, merry, pippin
The last two examples show recursive expansion with order-preserving
duplicate removal (hey there's now a published conference paper on
gitolite, so we have to use jargon *somewhere* or they won't accept
follow-on papers!).
If you do something like this:
config gitolite.mirror.nightly = "gimli gitolite.mirror.nightly"
or this:
config gitolite.mirror.nightly = "gimli gitolite.mirror.hourly"
config gitolite.mirror.hourly = "legolas gitolite.mirror.nightly"
you deserve what you get.
2. If you want to start a **foreground** job, the syntax is `gl-mirror-shell
request-push ip1 -fg gollum`. Foreground mode requires one (and only one)
slave name -- you cannot send to an implicit list, nor to more than one
slave.
3. Cronjobs and custom mirroring schemes are now very easy to do. Use either
of the command forms above and write a script around it. Appendix A
contains an example setup.
4. Once in a while a slave will realise it needs an update, and wants to ask
for one. It can run this command to do so:
ssh sam request-push ip2
If the requesting server is not one of the slaves listed in the config
variable gitolite.mirror.slaves on the master, it will be rejected.
This is always a foreground push, reflecting the fact that the slave may
want to know why their push errored out or didn't work last time or
whatever.
<a name="_details"></a>
### details
<a name="_the_conf_gitolite_conf_file"></a>
#### the `conf/gitolite.conf` file
One goal I have is to minimise the code changes to "core" gitolite due to
this, so all repo-specific mirror settings are stored as `git config`
variables (you know you can specify git config variables in the gitolite
config file right?). These are:
* `gitolite.mirror.master`
The name of the server which is the master for this repo. Each server
will compare this with `$GL_HOSTNAME` (from its own rc file) to
determine if it's the master or a slave. Here're the possible values:
* **undefined** or `local`: this repo is local to this server
* **same** as `$GL_HOSTNAME`: this server is the "master" for this
repo. (The repo is "native" to this server).
* **not same** as `$GL_HOSTNAME`: this server is a "slave" for the
repo. (The repo is a non-native on this server).
* `gitolite.mirror.slaves`
Ignored for non-native repos. For native repos, this is a space-separated
list of servers to push to from the `post-receive` hook.
Clearly, you can have different sets of slaves for different repos (again,
see "discussion" section later for more on this).
* `gitolite.mirror.redirectOK`
See the section on "redirecting pushes"
* In addition, you can create your own slave lists, named whatever you want,
except they have to start with `gitolite.mirror.`. The section on
"commands to (re-)sync mirrors" has some examples.
<a name="_redirecting_pushes"></a>
### redirecting pushes
**Please read carefully; there are security implications if you enable this
for mirrors NOT under your control**.
When a user pushes to a non-native repo, it is possible to transparently
redirect the push to the correct master server. This is a very neat feature,
because now all your users just use one URL (the mirror nearest to them).
They don't need to know where the actual master is, and more importantly, if
you and the other admins change it, they don't need to know it changed!
The `gitolite.mirror.redirectOK` config variable decides where this
redirection is OK. If it is set to 'true', any valid 'slave' can redirect an
incoming non-native push from a developer. Otherwise, it contains a list of
slaves that are permitted to redirect pushes (this might happen if you don't
trust some of your slaves enough to accept a redirected push from them).
This check needs to pass on both the master and slave servers; both have a say
in deciding if this is allowed. (The master may have real reasons not to
allow this; see below. I cannot think of any real reason for the *slave* to
disable this, but it's there in case some admin doesn't like it).
There are some potential issues that you MUST consider before enabling this:
* (security) If the slave and master server are so different or autonomous
that a user, say "alice", on the slave is not guaranteed to be the same
one as "alice" on the master, then the master admin should NOT enable this
feature.
This is because, in this scheme, authentication happens on the slave, but
authorisation is on the master. The slave-authenticated userid (alice) is
passed to the master.
(If you know ssh well enough, you know that the ssh authentication has
already happened, so all we can do is ensure authorisation happens with
whatever username we know so far).
* If your slave is out of sync with the master for whatever reason, then the
user will get confusing results. A `git fetch` may say everything is
upto-date but the push fails saying it is not a fast-forward push. (Of
course there's a way to fix this; see the "commands to (re-)sync mirrors"
section above).
* We cannot redirect non-git commands like ADC, setperms, etc because we
don't really have a way of knowing what repo he's talking about (different
commands have different syntaxes, some have more than one reponame...).
Any user who needs to do that should access the end server directly. It
should be easy enough to write an ADC to do the forwarding, in case the
slave server is the only one that can reach the real master due to network
or firewall setup.
Ideally, I recommend that ad hoc repos not be mirrored at all. Keep
mirroring for "blessed" repos only.
<a name="_discussion"></a>
### discussion
<a name="_problems_with_the_old_mirroring_model"></a>
#### problems with the old mirroring model
The old mirroring model had a single server as the master for *all*
repositories. Slaves were effectively only for load-balancing reads, or for
failover if the master died.
This is not good enough for corporate setups where the developers are spread
fairly evenly across the world. Some repos need to be closer to some teams
(NUMA is a good analogy).
A model where different repos are "mastered" in different cities is much more
efficient here.
The old model had other rigidities too, though they're not really *problems*,
as such:
* the slaves are just slaves; they can't have any "local" repos.
* a slave had to carry *all* repos; it couldn't choose to carry just a
subset.
* it implicitly assumed all the mirrors were under the same admin, and that
the gitolite-admin repo was itself mirrored too.
<a name="_the_new_mirroring_model"></a>
#### the new mirroring model
In the new model, servers can be much more independent and autonomous than in
the old model. (Don't miss the side note in the 'repository level setup'
section if you prefer the old model).
The new model has a few pros and cons. The pros come from the flexibility and
freedom that mirrors servers get, and the cons come from authorisation being
more rigorously checked (for example, a slave will only accept a push if *its*
configuration also says that the sending server is indeed the master for this
repo).
* A mirroring operation will not *create* a repo on the mirror; it has to
exist before a push happens on the master. Typically, the admin on the
slave must create the repo by adding the appropriate lines in his config.
If your setup is not autonomous (i.e., you're mirroring the admin repo as
well) then this happens automatically for normal repos. However,
*wildcard repos still won't work as seamlessly as in the old model*; see
the first bullet in the 'IMPORTANT cautions' section earlier.
* The gitolite-admin repo (and config) need not be mirrored. This allows
the slave server admin to create site-local repos, without forcing him to
create a second gitolite install for them.
(Site-local repos are useful for purely local projects that need
not/should not be mirrored for some reason, or ad-hoc personal repos that
developers create for themselves, etc.)
* Servers can choose to mirror a subset of the repos from one of the bigger
servers.
In the open source world, you can imagine more popular repos (or more
popular parts of huge projects like KDE) having more mirrors. Or
substitute "more popular" with "larger in size" if you wish
(FlightGear-data anyone?)
In the corporate world it could help with jurisdiction issues if the
mirror is in a different country with different laws.
I'm sure people will find other uses for this. And I'm *positive* the
pros will outweigh the cons. If you don't like it, follow the suggestion
in the side note somewhere up above, and just forget this feature exists
:-)
----
<a name="_appendix_A_example_cronjob_based_mirroring"></a>
### appendix A: example cronjob based mirroring
Let's say you have some repos that are very active. You're pushing halfway
across the world every few seconds, but those slaves do not need to be that closely
updated, perhaps *because* they are halfway across the world and those guys
are asleep ;-)
You'd like to update them once an hour instead. Here's how you might do that.
First add this line to the configuration for those repos:
config gitolite.mirror.hourly = "slave1 slave2 slave3"
Then write a cron job that looks like this (untested).
#!/bin/bash
REPO_BASE=`${0%/*}/gl-query-rc REPO_BASE`
cd $REPO_BASE
find . -type d -name "*.git" -prune | while read r
do
# get reponame as gitolite knows it
r=${r:2}
r=${r%.git}
gl-mirror-shell request-push $r gitolite.mirror.hourly
# that command backgrounds the push, so you'd best wait a few seconds
# before hitting the next one, otherwise you'll have all your repos
# going out at once!
sleep 10
done
<a name="_appendix_B_efficiency_versus_paranoia"></a>
### appendix B: efficiency versus paranoia
If you're paranoid enough to use mirrors, you should be paranoid enough to
like the `receive.fsckObjects` setting we now default to :-) However, informal
tests indicate a 40-50% CPU overhead from this. If you don't like that,
remove that line from the post-receive code.
use the `receive.fsckObjects` setting. However, informal tests indicate a
40-50% CPU overhead from this. If you're ok with that, make the appropriate
adjustments to `GL_GITCONFIG_KEYS` in the rc file, then add this to your
gitolite.conf file:
Please also note that we only set it on mirrors, and that too at the time the
mirrored repo is *created*. This means, when you start using your old "main"
server as a mirror (see later sections on switching over to a mirror, etc.),
it's repos do not have this setting. Repos created by previous versions of
gitolite also will not have this setting.
repo @all
config receive.fsckObjects = "true"
Personally, I just set `git config --global receive.fsckObjects true`, since
those servers aren't doing anything else anyway, and are idle for long
stretches of time. It's upto you what you want to do here.
<a name="_syncing_the_mirrors_the_first_time"></a>
[ch]: http://sitaramc.github.com/gitolite/doc/2-admin.html#_custom_hooks
[ha]: http://sitaramc.github.com/gitolite/doc/ssh-troubleshooting.html#_appendix_4_host_aliases
[rsgc]: http://sitaramc.github.com/gitolite/doc/gitolite.conf.html#_repo_specific_git_config_commands
[vsi]: http://sitaramc.github.com/gitolite/doc/gitolite.rc.html#_variables_with_a_security_impact
### syncing the mirrors the first time
This is fine if you're setting up everything from scratch. But if your master
server already had some repos with commits on them, you have to manually sync
them up once.
# on foo
gl-mirror-sync gitolite@bar
# path to "sync" program is ~/.gitolite/src if "from-client" install
<a name="_switching_over"></a>
### switching over
Let's say foo goes down. You want to make bar the main server, and continue
to have "baz" be a slave.
* on bar, edit `~/.gitolite.rc` and set
$GL_SLAVE_MODE = 0;
$ENV{GL_SLAVES} = 'gitolite@baz';
* **sanity check**: go to your gitolite-admin clone, add a remote for "bar",
fetch it, and make sure they are the same:
git remote add bar gitolite@bar:gitolite-admin
git fetch bar
git branch -a -v
# check that all SHAs are the same
* inform everyone of the new URL for their repos (see next section for more
on this)
* make sure that if "foo" does come up, it will not immediately start
serving requests. You'll be in trouble if (a) foo comes up as it was
before, and (b) some developer still had the old URL lying around and
started pushing changes to it.
You could jump in quickly and set `$GL_SLAVE_MODE = 1` as soon as the
system comes up. Better still, use extraneous means to block incoming
connections from normal users (out of scope for this document).
<a name="_the_return_of_foo"></a>
### the return of foo
<a name="_switching_back"></a>
#### switching back
Switching back is fairly easy.
* synchronise all repos from bar to foo. This may take some time, depending
on how long foo was down.
# on bar
gl-mirror-sync gitolite@foo
# path to "sync" program is ~/.gitolite/src if "from-client" install
* turn off pushes on "bar" by setting slave mode to 1
* run the sync once again; this should complete quickly
* **double check by comparing some the repos on both sides if needed**. You
could run the following snippet on all servers for a quick check:
cd ~/repositories # or wherever $REPO_BASE is
find . -type d -name "*.git" | sort |
while read r
do
echo $r
git ls-remote $r | sort
done | md5sum
* on foo, set the slave list (or check that it is correct)
* on foo, set slave mode off
* tell everyone to switch back
<a name="_making_foo_a_slave"></a>
#### making foo a slave
If "foo" does come up in a controlled manner, you might not want to switch
back right away. Unless you're doing DNS tricks, users may be peeved at
having to do 2 switches.
If you want to make foo a slave, you know the drill by now:
* set slave mode to 1 on foo
* on bar, add foo as a slave
# in ~/.gitolite.rc on bar
$ENV{GL_SLAVES} = 'gitolite@foo gitolite@baz';
I think that should cover pretty much everything. I *have* tested most of
this, but YMMV.
----
<a name="_URLs_that_your_users_will_use"></a>
### URLs that your users will use
Unless you play DNS tricks, it is more than likely that your users would have
to change the URLs they use to access their repos if you change the server
they push to.
I cannot speak for the plethora of git client software out there but for
normal git, this problem can be mitigated somewhat by doing this:
* in `~/.ssh/config` on my workstation, I have
host gl
hostname=primary.server.ip
user=gitolite
* all my `git clone` commands use `gl:reponame` as the URL
* if the primary goes down, and I have to access the secondary, I just
change the `hostname` line in `~/.ssh/config`.
That's it. Every clone of every repo used anywhere in this userid is now
changed.
To repeat, this may or may not work with all the git clients that exist (like
jgit, or any of the GUI tools, and especially if you're on Windows).
If anyone has a better idea, something that works more universally, I'd love
to hear it.

View file

@ -8,19 +8,17 @@
# if you don't do this, git-shell sometimes dies of a signal 13 (SIGPIPE)
[ -t 0 ] || cat >/dev/null
if [ -n "$GL_SLAVES" ]
then
for mirror in $GL_SLAVES
do
if git push --mirror $mirror:$GL_REPO.git
then
:
else
ssh $mirror mkdir -p $GL_REPO.git
ssh $mirror git init --bare $GL_REPO.git
ssh $mirror "cd $GL_REPO.git; git config receive.fsckObjects true"
git push --mirror $mirror:$GL_REPO.git ||
echo "WARNING: mirror push to $mirror failed"
fi
done
fi >&2
# even slaves have post-receive hooks, but due to the way the push happens, we
# don't have GL_REPO set. So we detect that generic situation and bail...
[ -n "$GL_BYPASS_UPDATE_HOOK" ] && exit 0
# CAUTION: this means that a server-side push (bypassing gitolite) will not be
# mirrored automatically because (a) we don't know GL_REPO (we can deduce it
# but we won't!), and (b) we can't distinguish easily between that and this
# case (the slave receiving a mirror push case)
[ -z "$GL_REPO" ] && die GL_REPO not set
[ -z "$GL_BINDIR" ] && die GL_BINDIR not set
slaves=`git config --get gitolite.mirror.slaves`
[ -z "$slaves" ] && exit 0
$GL_BINDIR/gl-mirror-push $GL_REPO $slaves

View file

@ -22,11 +22,16 @@ use Exporter 'import';
setup_git_configs
setup_gitweb_access
shell_out
slurp
special_cmd
try_adc
wrap_chdir
wrap_open
wrap_print
mirror_mode
mirror_listslaves
mirror_redirectOK
);
@EXPORT_OK = qw(
%repos
@ -159,7 +164,8 @@ sub log_it {
$logmsg .= "\t@_" if @_;
# erm... this is hard to explain so just see the commit message ok?
$logmsg =~ s/([\x00-\x08\x0B\x0C\x0E-\x1F\x7F-\xFF]+)/sprintf "<<hex(%*v02X)>>","",$1/ge;
print $log_fh "$ENV{GL_TS}\t$ENV{GL_USER}\t$ip\t$logmsg\n";
my $user = $ENV{GL_USER} || "(no user)";
print $log_fh "$ENV{GL_TS}\t$user\t$ip\t$logmsg\n";
close $log_fh or die "close log failed: $!\n";
}
@ -420,12 +426,24 @@ sub setup_git_configs
{
my ($repo, $git_configs_p) = @_;
while ( my ($key, $value) = each(%{ $git_configs_p->{$repo} }) ) {
if ($value ne "") {
$value =~ s/^"(.*)"$/$1/;
system("git", "config", $key, $value);
} else {
system("git", "config", "--unset-all", $key);
# new_wild calls us without checking!
return unless $git_configs_p->{$repo};
# git_configs_p is a ref to a hash whose elements look like
# {"reponame"}{sequence_number}{"key"} = "value";
my %rch = %{ $git_configs_p->{$repo} };
# %rch has elements that look like {sequence_number}{"key"} = "value"
for my $seq (sort { $a <=> $b } keys %rch) {
# and the final step is the repo config: {"key"} = "value"
my $rc = $rch{$seq};
while ( my ($key, $value) = each(%{ $rc }) ) {
if ($value ne "") {
$value =~ s/^"(.*)"$/$1/;
system("git", "config", $key, $value);
} else {
system("git", "config", "--unset-all", $key);
}
}
}
}
@ -1182,6 +1200,49 @@ sub ext_cmd_svnserve
die "svnserve exec failed\n";
}
# ----------------------------------------------------------------------------
# MIRRORING HELPERS
# ----------------------------------------------------------------------------
sub mirror_mode {
my $repo = shift;
# 'local' is the default if the config is empty or not set
my $gmm = `git config --file $REPO_BASE/$repo.git/config --get gitolite.mirror.master` || 'local';
chomp $gmm;
return 'local' if $gmm eq 'local';
return 'master' if $gmm eq ( $GL_HOSTNAME || '' );
return "slave of $gmm";
}
sub mirror_listslaves {
my $repo = shift;
return ( `git config --file $REPO_BASE/$repo.git/config --get gitolite.mirror.slaves` || '' );
}
# is a redirect ok for this repo from this slave?
sub mirror_redirectOK {
my $repo = shift;
my $slave = shift || return 0;
# if we don't know who's asking, the answer is "no"
my $gmrOK = `git config --file $REPO_BASE/$repo.git/config --get gitolite.mirror.redirectOK` || '';
chomp $gmrOK;
my $slavelist = mirror_listslaves($repo);
# if gmrOK is 'true', any valid slave can redirect
return 1 if $gmrOK eq 'true' and $slavelist =~ /(^|\s)$slave(\s|$)/;
# otherwise, gmrOK is a list of slaves who can redirect
return 1 if $gmrOK =~ /(^|\s)$slave(\s|$)/;
return 0;
# LATER/NEVER: include a call to an external program to override a 'true',
# based on, say, the time of day or network load etc. Cons: shelling out,
# deciding the name of the program (yet another rc var?)
}
# ------------------------------------------------------------------------------
# per perl rules, this should be the last line in such a file:
1;

View file

@ -17,12 +17,13 @@ use Exporter 'import';
$ADMIN_POST_UPDATE_CHAINS_TO $ENV $GITOLITE_BASE $GITOLITE_PATH $GIT_PATH
$GL_ADC_PATH $GL_ADMINDIR $GL_ALL_INCLUDES_SPECIAL $GL_ALL_READ_ALL
$GL_BIG_CONFIG $GL_CONF $GL_CONF_COMPILED $GL_GET_MEMBERSHIPS_PGM
$GL_GITCONFIG_KEYS $GL_GITCONFIG_WILD $GL_KEYDIR $GL_LOGT $GL_NICE_VALUE
$GL_GITCONFIG_KEYS $GL_KEYDIR $GL_LOGT $GL_NICE_VALUE
$GL_NO_CREATE_REPOS $GL_NO_DAEMON_NO_GITWEB $GL_NO_SETUP_AUTHKEYS
$GL_PACKAGE_CONF $GL_PACKAGE_HOOKS $GL_PERFLOGT $GL_SITE_INFO
$GL_SLAVE_MODE $GL_WILDREPOS $GL_WILDREPOS_DEFPERMS
$GL_WILDREPOS_PERM_CATS $HTPASSWD_FILE $PROJECTS_LIST $REPO_BASE
$REPO_UMASK $RSYNC_BASE $SVNSERVE $UPDATE_CHAINS_TO $AUTH_OPTIONS
$GL_HOSTNAME
$GL_HTTP_ANON_USER
);
@ -31,7 +32,7 @@ use Exporter 'import';
# real constants
# ------------------------------------------------------------------------------
$current_data_version = '1.7';
$current_data_version = '2.0';
$ABRT = "\n\t\t***** ABORTING *****\n ";
$WARN = "\n\t\t***** WARNING *****\n ";
@ -72,6 +73,11 @@ do $ENV{GL_RC} or die "error parsing $ENV{GL_RC}\n";
# fix up REPO_BASE
$REPO_BASE = "$ENV{HOME}/$REPO_BASE" unless $REPO_BASE =~ m(^/);
# <sigh> backward incompat detection for mirroring. Normally I wouldn't do
# this but this is *important*
die "$ABRT Mirroring has completely changed in this version.\tYou need to check the documentation for how to upgrade\n"
if (defined $GL_SLAVE_MODE or exists $ENV{GL_SLAVES});
# ------------------------------------------------------------------------------
# per perl rules, this should be the last line in such a file:
1;

View file

@ -93,10 +93,6 @@ unless ($ENV{SSH_ORIGINAL_COMMAND}) {
$ENV{SSH_ORIGINAL_COMMAND} = 'info';
}
# slave mode should not do much
die "server is in slave mode; you can only fetch\n"
if ($GL_SLAVE_MODE and $ENV{SSH_ORIGINAL_COMMAND} !~ /^(info|expand|get|git-upload-)/);
# admin defined commands; please see doc/admin-defined-commands.mkd
if ($GL_ADC_PATH and -d $GL_ADC_PATH) {
try_adc(); # if it succeeds, this also 'exec's out
@ -139,6 +135,19 @@ $ENV{GL_REPO}=$repo;
# the real git commands (git-receive-pack, etc...)
# ----------------------------------------------------------------------------
# we know the user and repo; we just need to know what perm he's trying for
# (aa == attempted access; setting this makes some later logic simpler)
my $aa = ($verb =~ $R_COMMANDS ? 'R' : 'W');
# writes may get redirected under certain conditions
if ( $aa eq 'W' and mirror_mode($repo) =~ /^slave of (\S+)/ ) {
my $master = $1;
die "$ABRT GL_HOSTNAME not set; rejecting push to non-local repo\n" unless $GL_HOSTNAME;
die "$ABRT $GL_HOSTNAME not the master, please push to $master\n" unless mirror_redirectOK($repo, $GL_HOSTNAME);
print STDERR "$GL_HOSTNAME ==== $user ($repo) ===> $master\n";
exec("ssh", $master, "USER=$user", "SOC=$ENV{SSH_ORIGINAL_COMMAND}");
}
# first level permissions check
my ($perm, $creator, $wild);
@ -150,9 +159,6 @@ if ( $GL_ALL_READ_ALL and $verb =~ $R_COMMANDS and -d "$REPO_BASE/$repo.git") {
# it was missing, and you have create perms, so create it
new_wild_repo($repo, $user) if ($perm =~ /C/);
# we know the user and repo; we just need to know what perm he's trying for
# (aa == attempted access)
my $aa = ($verb =~ $R_COMMANDS ? 'R' : 'W');
die "$aa access for $repo DENIED to $user
(Or there may be no repository at the given path. Did you spell it correctly?)\n" unless $perm =~ /$aa/;

View file

@ -49,8 +49,9 @@ open STDOUT, ">", "/dev/null" if (@ARGV and shift eq '-q');
# names of repos whose ACLs don't make it into the main compiled config file
# copy above desc to lite.pm -- my %split_conf = ();
# rule sequence number
# rule and config sequence numbers
my $rule_seq = 0;
my $config_seq = 0;
# <sigh>... having been forced to use a list as described above, we lose some
# efficiency due to the possibility of the same {ref, perms} pair showing up
@ -244,21 +245,11 @@ sub parse_conf_line
die "$ABRT git config $key not allowed\ncheck GL_GITCONFIG_KEYS in the rc file for how to allow it\n" if (@matched < 1);
for my $repo (@{ $repos_p }) # each repo in the current stanza
{
$git_configs{$repo}{$key} = $value;
$git_configs{$repo}{$config_seq++}{$key} = $value;
# force entry in %repos. Without this, a repo para with just a
# config line and no ACLs gets ignored in the output
$repos{$repo}{HAS_CONFIG} = 1;
# no problem if it's a plain repo (non-pattern, non-groupname)
# OR wild configs are allowed
unless ( ($repo =~ $REPONAME_PATT and $repo !~ /^@/) or $GL_GITCONFIG_WILD) {
my @r = ($repo); # single wildpatt
@r = sort keys %{ $groups{$repo} } if $groups{$repo}; # or a group; get its members
do {
warn "$WARN git config set for $_ but \$GL_GITCONFIG_WILD not set\n" unless $_ =~ $REPONAME_PATT
} for @r;
}
}
}
# include

83
src/gl-mirror-push Executable file
View file

@ -0,0 +1,83 @@
#!/bin/sh
# arguments: reponame, list of slaves
# optional flag after reponame: "-fg" to run in foreground. This is only
# going to be given by one specific invocation, and if given will only work
# for one slave.
# if list of slaves not given, get it from '...slaves' config
die() { echo gl-mirror-push${hn:+ on $hn}: "$@" >&2; exit 1; }
get_rc_val() { ${0%/*}/gl-query-rc $1; }
# ----------
# is mirroring even enabled?
hn=`get_rc_val GL_HOSTNAME`
[ -z "$hn" ] && exit
# we should not be invoked directly from the command line
[ -z "$GL_LOG" ] && die fatal: do not run $0 directly
# ----------
# get repo name then check if it's a local or slave (ie we're not the master)
[ -z "$1" ] && die fatal: missing reponame argument
repo=$1; shift
REPO_BASE=`get_rc_val REPO_BASE`
cd $REPO_BASE/$repo.git 2>/dev/null || die fatal: could not change directory to "$repo"
gmm=`git config --get gitolite.mirror.master`
# is it local? (remember, empty/undef ==> local
gmm=${gmm:-local}
[ "$gmm" = "local" ] && exit
# is it a slave?
[ "$hn" = "$gmm" ] || die fatal: wrong master. Try $gmm...
# ----------
# now see if we want to be foregrounded. Fg mode accepts only one slave
[ "$1" = "-fg" ] && {
[ -z "$2" ] && die fatal: missing slavename argument
[ -n "$3" ] && die fatal: too many slavenames
git push --mirror $2:$repo 2>&1 | sed -e "s/^/$hn:/"
exit
}
# ----------
# normal (self-backgrounding) mode, one or more slaves
[ -z "$1" ] && die fatal: missing list of slaves
export slaves
slaves="$*"
# ----------
# print out the job ID, then redirect all 3 FDs
export job_id=$$ # can change to something else if needed
echo "($job_id&) $hn ==== ($repo) ===>" $slaves >&2
logfile=${GL_LOG/%.log/-mirror-pushes.log}
exec >>$logfile 2>&1 </dev/null
# ----------
# and finally...
# double fork, no zombies
(
(
echo `date +%T` $repo '===>'
for s in $slaves
do
[ "$s" = "$hn" ] && continue # skip ourselves
git push --mirror $s:$repo || echo ==== WARNING: RC=$? from git push --mirror $s:$repo ====
done 2>&1 | sed -e "s/^/ /"
echo `date +%T` '===>' $slaves
echo
) 2>&1 | sed -e "s/^/$job_id:/" & # background the whole thing
)

View file

@ -1,30 +1,168 @@
#!/bin/bash
#!/usr/bin/perl
export GL_BYPASS_UPDATE_HOOK
GL_BYPASS_UPDATE_HOOK=1
# terminology:
# native repo: a repo for which we are the master; pushes happen here
# authkeys: shorthand for ~/.ssh/authorized_keys
get_rc_val() {
${0%/*}/gl-query-rc $1
# this is invoked in one of two ways:
# (1) locally, from a shell script or command line
# (2) from a remote server, via authkeys, with one argument (the name of the
# sending server), similar to what happens with normal users and the
# 'gl-auth-command' program. SSH_ORIGINAL_COMMAND will then contain the
# actual command that the remote sent.
#
# Currently, these commands are (a) 'info', (b) 'git-receive-pack' when a
# mirror push is *received* by a slave, (c) 'request-push' sent by a slave
# (possibly via an ADC) when the slave finds itself out of sync, (d) a
# redirected push, from a user pushing to a slave, which is represented not by
# a command per se but by starting with "USER=..."
use strict;
use warnings;
# ----------------------------------------------------------------------------
# this section of code snarfed from gl-auth-command
# XXX add this program to 'that bindir thing' in doc/developer-notes.mkd
BEGIN {
$0 =~ m|^(/)?(.*)/| and $ENV{GL_BINDIR} = ($1 || "$ENV{PWD}/") . $2;
}
REPO_BASE=$( get_rc_val REPO_BASE)
REPO_UMASK=$(get_rc_val REPO_UMASK)
use lib $ENV{GL_BINDIR};
umask $REPO_UMASK
use gitolite_rc;
use gitolite_env;
use gitolite;
if echo $SSH_ORIGINAL_COMMAND | egrep git-upload\|git-receive >/dev/null
then
setup_environment();
die "fatal: GL_HOSTNAME not set in rc; mirroring disabled\n" unless $GL_HOSTNAME;
# the (special) admin post-update hook needs these, so we cheat
export GL_RC
export GL_ADMINDIR
export GL_BINDIR
GL_RC=$( get_rc_val GL_RC)
GL_ADMINDIR=$(get_rc_val GL_ADMINDIR)
GL_BINDIR=$( get_rc_val GL_BINDIR)
# ----------------------------------------------------------------------------
SSH_ORIGINAL_COMMAND=`echo $SSH_ORIGINAL_COMMAND | sed -e "s:':'$REPO_BASE/:"`
exec git shell -c "$SSH_ORIGINAL_COMMAND"
else
bash -c "cd $REPO_BASE; $SSH_ORIGINAL_COMMAND"
fi
# deal with local invocations first
# on the "master", run from a shell, for one specific repo, with an optional
# list of slaves, like so:
# gl-mirror-shell request-push some-repo [optional list of slaves/keys]
if ( ($ARGV[0] || '') eq 'request-push' and not $ENV{SSH_ORIGINAL_COMMAND} ) {
shift;
my $repo = shift or die "fatal: missing reponame\n";
-d "$REPO_BASE/$repo.git" or die "fatal: no such repo?\n";
# this is the default argument if no slave list or key is supplied
@ARGV = ('gitolite.mirror.slaves') unless @ARGV;
my @slaves = ();
my %seen = ();
# each argument in @ARGV is either a slave name, or a gitolite mirroring
# key to be replaced with its value, split into a list of slaves
while (@ARGV) {
$a = shift @ARGV;
if ($a =~ /^gitolite\.mirror\.[\w.-]+$/) {
my @values = split(' ', `git config --file $REPO_BASE/$repo.git/config --get $a` || '');
unshift @ARGV, @values;
} else {
push @slaves, $a unless $seen{$a}++;
}
}
exit 1 unless @slaves;
# we don't want to complain louder than that because the most common
# use of this script on the master server is via cron, run against
# *all* known repos without checking their individual key values
print STDERR "info: mirror-push $repo ", join(" ", @slaves), "\n";
system("gl-mirror-push", $repo, @slaves);
exit 0;
}
unless (@ARGV) { print STDERR "fatal: missing command\n"; exit 1; }
# ----------
# now the remote invocations; log it, then get the sender name
my $sender = shift;
$ENV{GL_USER} ||= "host:$sender";
# default SSH_ORIGINAL_COMMAND is 'info', as usual
$ENV{SSH_ORIGINAL_COMMAND} ||= 'info';
# and it's too long to bloody type...
my $soc = $ENV{SSH_ORIGINAL_COMMAND};
log_it();
# ----------
# our famous 'info' command
if ($soc eq 'info') {
print STDERR "Hello $sender, I am $GL_HOSTNAME\n";
exit;
}
# ----------
# when running on the "slave", we have to "receive" the `git push --mirror`
# from a master. Check that the repo is indeed a slave and the sender is the
# correct master before allowing the push.
if ($soc =~ /^git-receive-pack '(\S+)'$/) {
my $repo = $1;
die "fatal: invalid characters in $repo\n" unless $repo =~ $REPONAME_PATT;
my $mm = mirror_mode($repo);
# reminder: we're not going through the slave-side gl-auth-command. This
# is a server-to-server transaction, with an authenticated sender.
# Authorisation consists of checking to make sure our config says this
# sender is indeed the master for this repo
die "$ABRT fatal: $GL_HOSTNAME <==//== $sender mirror-push rejected: $repo is $mm\n" unless $mm eq "slave of $sender";
print STDERR "$GL_HOSTNAME <=== ($repo) ==== $sender\n";
$ENV{GL_BYPASS_UPDATE_HOOK} = 1;
# replace the repo path with the full path and hand off to git-shell
$soc =~ s(')('$ENV{GL_REPO_BASE_ABS}/);
exec("git", "shell", "-c", $soc);
}
# ----------
# a slave may have found itself out of sync (perhaps the network was down at
# the time of the last push to the master), and now wants to request a sync.
# This is similar to the "local invocation" described above, but we check the
# sender name against gitolite.mirror.slaves to prevent some random slave from
# asking for a repo it should not be having!
if ($soc =~ /^request-push (\S+)$/) {
my $repo = $1;
die "fatal: invalid characters in $repo\n" unless $repo =~ $REPONAME_PATT;
die "$ABRT fatal: $GL_HOSTNAME ==//==> $sender refused: not in slave list\n" unless mirror_listslaves($repo) =~ /(^|\s)$sender(\s|$)/;
print STDERR "$GL_HOSTNAME ==== ($repo) ===> $sender\n";
# just one sender, and we've checked that he is "on the list". Foreground...
system("$ENV{GL_BINDIR}/gl-mirror-push", $repo, "-fg", $sender);
exit;
}
# ----------
# experimental feature...
# when running on the "master", receive a redirected push from a slave. This
# is disabled by default and needs to be explicitly enabled on both the master
# and the slave. SEE DOCUMENTATION FOR CAVEATS AND CAUTIONS.
if ($soc =~ /^USER=(\S+) SOC=(git-receive-pack '(\S+)')$/) {
my $user = $1;
$ENV{SSH_ORIGINAL_COMMAND} = $2;
my $repo = $3;
die "fatal: invalid characters in $user\n" unless $user =~ $USERNAME_PATT;
die "fatal: invalid characters in $repo\n" unless $repo =~ $REPONAME_PATT;
die "$ABRT fatal: $GL_HOSTNAME <==//== $sender redirected push rejected\n" unless mirror_redirectOK($repo, $sender);
print STDERR "$GL_HOSTNAME <=== $user ($repo) ==== $sender\n";
my $pgm = $0;
$pgm =~ s([^/]+$)(gl-auth-command);
exec($pgm, $user);
}

View file

@ -1,38 +0,0 @@
#!/bin/bash
mirror=$1
[ -z "$1" ] && { echo need \"user@host\" or ssh hostalias; exit 1; }
ssh -o PasswordAuthentication=no $mirror echo hello-there | grep hello-there >/dev/null ||
{ echo I cant ssh to $mirror; exit 1; }
cd $HOME
REPO_BASE=`${0%/*}/gl-query-rc REPO_BASE`
cd $REPO_BASE
ssh $mirror cat \$HOME/.gitolite.rc | expand | egrep '^ *\$GL_SLAVE_MODE *= *1; *$' >/dev/null || {
echo $mirror does not seem to be in slave mode
exit 1;
}
find . -type d -name "*.git" -prune | cut -c3- | sort | while read r
do
cd $HOME; cd $REPO_BASE; cd $r
printf "$r "
if [ `git rev-parse HEAD` = "HEAD" ]
then
echo is empty\; skipping
continue
fi
# this is essentially the same code as in the post-receive hook
if git push --mirror $mirror:$r
then
:
else
ssh $mirror mkdir -p $r
ssh $mirror git init --bare $r
git push --mirror $mirror:$r ||
echo "WARNING: mirror push to $mirror failed"
fi < /dev/null
done

View file

@ -1,75 +1,106 @@
#!/bin/sh
#!/usr/bin/perl -w
# BEGIN USAGE
# help/instructions are at the bottom, in the __DATA__ section
# $0 -- make some server side tasks easier
use strict;
use warnings;
# Usage:
# $0 [sub-command [args]]
use FindBin;
BEGIN { $ENV{GL_BINDIR} = $FindBin::Bin; }
# Security notes: this program does not do any sanitisation of input. You're
# running it at the CLI on the server, so you already have the power to do
# whatever you want anyway.
use lib $ENV{GL_BINDIR};
use gitolite_rc;
use gitolite;
# current sub-commands:
sub usage { print <DATA>; exit 1; }
usage() unless (@ARGV);
# (1) REPLACE THE OLD $SHELL_USERS MECHANISM
#
# $0 shell-add foo.pub
#
# adds the pubkey in foo.pub into the authkeys file with "-s" argument (shell
# access) and user "foo". The line will be added *before* the "# gitolite
# start" section, so that a gitolite-admin push will not affect it.
my $cmd = shift;
my $pub = shift;
# Although there is no "shell-remove" sub-command, you can do that quite
# easily by editing ~/.ssh/authorized_keys and deleting the appropriate line.
# END USAGE
die() { echo "$@"; exit 1; } >&2
if [ -z "$1" ]
then
perl -ne 's/\$0/$ARGV/ge; print if /BEGIN USAGE/../END USAGE/' $0 | grep -v USAGE | cut -c3-
exit 1
fi
if [ "$1" = "shell-add" ]
then
if ($cmd eq 'add-shell-user' or $cmd eq 'add-mirroring-peer') {
# sanity checks
[ -z "$2" ] && exec $0
[ -f "$2" ] || die "$2 does not exist"
wc -l < $2 | grep '^1$' >/dev/null || die "$2 contains more than one line"
$pub or usage();
my $user = validate_pubkeyfile($pub);
# write the file out, with the new authkeys line added just *before* the
# gitolite section. But first, set the command that gets invoked
$cmd = ( $cmd eq 'add-shell-user' ? 'gl-auth-command -s' : 'gl-mirror-shell' );
ak_insert($cmd, $user, $pub);
exit 0;
}
die "could not understand command $cmd\n";
sub validate_pubkeyfile {
my $pub = shift;
-f $pub or die "$pub does not exist\n";
die "$pub contains more than one line\n" if wc_l($pub) > 1;
my $user = $pub;
$user =~ s(^.*/)(); # remove optional directory
die "file name must end in .pub\n" unless $user =~ /(.*)\.pub$/;
$user = $1;
return $user;
}
sub ak_insert {
my ($cmd, $user, $pub) = @_;
# must be kept consistent with what's in src/gl-compile-conf; on the plus
# side, it's not likely to change anytime soon!
AUTH_OPTIONS="no-port-forwarding,no-X11-forwarding,no-agent-forwarding"
my $AUTH_OPTIONS = "no-port-forwarding,no-X11-forwarding,no-agent-forwarding";
GL_BINDIR=`${0%/*}/gl-query-rc GL_BINDIR`
my $authline = "command=\"$ENV{GL_BINDIR}/$cmd $user\",$AUTH_OPTIONS " . slurp($pub);
pubkey_file=$2
user=`basename $pubkey_file .pub`
my $authkeys = "$ENV{HOME}/.ssh/authorized_keys";
my $ak_lines = slurp($authkeys);
$ak_lines =~ s/^.*$cmd $user.*\n//m; # remove existing keyline, if present
$ak_lines =~ s/^# gitolite start/$authline# gitolite start/m;
my $akfh = wrap_open(">", $authkeys);
print $akfh $ak_lines;
close $akfh;
}
authline="command=\"$GL_BINDIR/gl-auth-command -s $user\",$AUTH_OPTIONS `cat $pubkey_file`";
sub wc_l {
my $fh = wrap_open("<", shift);
my @l = <$fh>;
my $l = @l;
return $l;
}
authkeys=$HOME/.ssh/authorized_keys
__DATA__
for i in 1
do
perl -lne "last if /# gitolite start/; print unless /gl-auth-command -s $user/; " $authkeys
echo $authline
perl -lne "print if /# gitolite start/ .. 0; " $authkeys
done > $authkeys.new
gl-tool -- make some server side tasks easier
diff -u $authkeys $authkeys.new && die no change to authkey file
echo
echo If the above diff looks ok, press enter. Else press Ctrl-C.
read dummy
cat $authkeys > $authkeys.old
cat $authkeys.new > $authkeys
Usage:
gl-tool [sub-command [args]]
exit 0
fi
Security notes: this program does not do any sanitisation of input. You're
running it at the CLI on the server, so you already have the power to do
whatever you want anyway.
die "could not understand command $1"
current sub-commands:
(1) REPLACE THE OLD $SHELL_USERS MECHANISM
gl-tool add-shell-user foo.pub
Adds the pubkey in foo.pub into the authkeys file with "-s" argument (shell
access) and user "foo". The line will be added *before* the "# gitolite
start" section, so that a gitolite-admin push will not affect it.
Although there is no "remove-shell-user" sub-command, you can do that quite
easily by editing ~/.ssh/authorized_keys and deleting the appropriate line.
(2) ADD A MIRRORING PEER KEY
gl-tool add-mirroring-peer git@server.company.com.pub
As above, but the given key will invoke 'gl-mirror-shell' instead of the
usual 'gl-auth-command'. This is meant to be a server-to-server key, allowing
(in this example), the gitolite server called 'git@server.company.com' to
access this server for mirroring operations.

View file

@ -1,4 +1,4 @@
$data_version = '1.7';
$data_version = '2.0';
%repos = (
'aa' => {
'R' => {

View file

@ -1,4 +1,4 @@
$data_version = '1.7';
$data_version = '2.0';
%repos = (
'aa' => {
'R' => {

View file

@ -1,4 +1,4 @@
$data_version = '1.7';
$data_version = '2.0';
%repos = ();
%split_conf = (
'aa' => 1,

View file

@ -1,4 +1,4 @@
$data_version = '1.7';
$data_version = '2.0';
%repos = (
'@g1' => {
'@g1' => [

View file

@ -1,4 +1,4 @@
$data_version = '1.7';
$data_version = '2.0';
%repos = (
'aa' => {
'R' => {

View file

@ -1,4 +1,4 @@
$data_version = '1.7';
$data_version = '2.0';
%repos = (
'aa' => {
'R' => {

View file

@ -1,4 +1,4 @@
$data_version = '1.7';
$data_version = '2.0';
%repos = ();
%split_conf = (
'aa' => 1,

View file

@ -1,4 +1,4 @@
$data_version = '1.7';
$data_version = '2.0';
%repos = (
'aa' => {
'@g1' => [

View file

@ -1,4 +1,4 @@
$data_version = '1.7';
$data_version = '2.0';
%repos = ();
%groups = (
'@g1' => {

View file

@ -1,7 +1,7 @@
# vim: syn=sh:
for bc in 0 1
do
for gcw in 0 1
for gcw in 0
do
cd $TESTDIR
$TESTDIR/rollback || die "rollback failed"
@ -9,7 +9,6 @@ do
name INTERNAL
editrc GL_WILDREPOS 1
editrc GL_BIG_CONFIG $bc
echo "\$GL_GITCONFIG_WILD = $gcw;" | addrc
# ----------
@ -47,8 +46,6 @@ do
RW = @leads
config foo.bar = baz
" | ugc -r
[ "$gcw" = "0" ] && expect "remote: git config set for bar/..\* but \$GL_GITCONFIG_WILD not set"
[ "$gcw" = "1" ] && notexpect "remote: git config set for bar/..\* but \$GL_GITCONFIG_WILD not set"
notexpect "git config.*not allowed"
expect_push_ok "master -> master"

View file

@ -1,12 +1,11 @@
# vim: syn=sh:
for gcw in 0 1
for gcw in 0
do
cd $TESTDIR
$TESTDIR/rollback || die "rollback failed"
name INTERNAL
editrc GL_WILDREPOS 1
echo "\$GL_GITCONFIG_WILD = $gcw;" | addrc
# ----------
@ -42,8 +41,6 @@ do
RW = @leads
config foo.bar = baz
" | ugc -r
[ "$gcw" = "0" ] && expect "remote: git config set for bar/\$creator/..\* but \$GL_GITCONFIG_WILD not set"
[ "$gcw" = "1" ] && notexpect "remote: git config set for bar/\$creator/..\* but \$GL_GITCONFIG_WILD not set"
notexpect "git config.*not allowed"
expect_push_ok "master -> master"