Changeset - 3389f272ece1
[Not reviewed]
beta
0 2 0
Marcin Kuzminski - 15 years ago 2010-11-23 22:05:30
marcin@python-works.com
docs update
2 files changed with 9 insertions and 3 deletions:
0 comments (0 inline, 0 general)
docs/index.rst
Show inline comments
 
.. _index:
 

	
 
Welcome to RhodeCode (RhodiumCode) documentation!
 
=================================================
 

	
 
``RhodeCode`` (formerly hg-app) is Pylons based Mercurial repository 
 
browser/management with build in push/pull server and full text search. It's 
 
similar to github or bitbucket, but it's suppose to run as standalone hosted 
 
application, it's open source and focuses more on restricted access to 
 
repositories. It's powered by vcs_ library that we created to handle many 
 
various version control systems.
 

	
 
RhodeCode uses `Semantic Versioning <http://semver.org/>`_
 

	
 
RhodeCode demo
 
--------------
 

	
 
http://hg.python-works.com
 

	
 
The default access is
 

	
 
- username: demo
 
- password: demo
 

	
 
Source code
 
-----------
 

	
 
Source code is along with issue tracker is available at
 
http://bitbucket.org/marcinkuzminski/rhodecode
 

	
 
Also a source codes can be obtained from demo RhodeCode instance
 
http://hg.python-works.com/rhodecode/summary
 

	
 
Features
 
--------
 

	
 
- Has it's own middleware to handle mercurial_ and git_ protocol request. 
 
  Each request can be logged and authenticated. Runs on threads unlikely to 
 
  hgweb You can make multiple pulls/pushes simultaneous. Supports http/https
 
  both on git_ and mercurial_
 
- Full permissions and authentication per project private/read/write/admin. 
 
  One account for web interface and mercurial_ push/pull/clone.
 
- Mako templates let's you customize look and feel of application.
 
- Beautiful diffs, annotations and source codes all colored by pygments.
 
- Mercurial_ branch graph and yui-flot powered graphs with zooming and statistics
 
- Admin interface with user/permission management. User activity journal logs
 
  pulls, pushes, forks,registrations. Possible to disable built in hooks
 
- Server side forks, it's possible to fork a project and hack it free without
 
  breaking the main.   
 
- Full text search on source codes, search on file names. All powered by whoosh
 
  and build in indexing daemons
 
  (no external search servers required all in one application)
 
- Rss / atom feeds, gravatar support, download sources as zip/tarballs  
 
- Async tasks for speed and performance using celery_ (works without them too)  
 
- Backup scripts can do backup of whole app and send it over scp to desired 
 
  location
 
- Setup project descriptions and info inside built in db for easy, non 
 
  file-system operations
 
- Added cache with invalidation on push/repo management for high performance and
 
  always up to date data. 
 
- Based on pylons 1.0 / sqlalchemy 0.6 / sqlite
 

	
 

	
 
.. figure::  images/screenshot1_main_page.png
 
   :align:   left
 

	
 
   Main page of RhodeCode
 

	
 
.. figure::  images/screenshot2_summary_page.png
 
   :align:   left
 

	
 
   Summary page
 

	
 
.. figure::  images/screenshot3_changelog_page.png
 

	
 
   Changelog with DAG graph
 
    
 

	
 
Incoming / Plans
 
----------------
 

	
 
- code review (probably based on hg-review)
 
- full git_ support, with push/pull server
 
- project grouping
 
- redmine integration
 
- commit based build in wiki system
 
- clone points and cloning from remote repositories into rhodecode 
 
  (git_ and mercurial_)
 
- more statistics and graph (global annotation + some more statistics)
 
- user customized activity dashboards
 
- other cools stuff that i can figure out (or You can help me figure out)
 

	
 
License
 
-------
 

	
 
``rhodecode`` is released under GPL_ license.
 

	
 

	
 
Mailing group Q&A
 
-----------------
 

	
 
http://groups.google.com/group/rhodecode
 

	
 

	
 
Documentation
 
-------------
 

	
 
**Installation:**
 

	
 
.. toctree::
 
   :maxdepth: 1
 

	
 
   installation
 
   setup
 
   upgrade
 
   setup
 
   changelog
 

	
 
Other topics
 
------------
 

	
 
* :ref:`genindex`
 
* :ref:`search`
 

	
 
.. _virtualenv: http://pypi.python.org/pypi/virtualenv
 
.. _python: http://www.python.org/
 
.. _django: http://www.djangoproject.com/
 
.. _mercurial: http://mercurial.selenic.com/
 
.. _subversion: http://subversion.tigris.org/
 
.. _git: http://git-scm.com/
 
.. _celery: http://celeryproject.org/
 
.. _Sphinx: http://sphinx.pocoo.org/
 
.. _GPL: http://www.gnu.org/licenses/gpl.html
 
.. _vcs: http://pypi.python.org/pypi/vcs
docs/setup.rst
Show inline comments
 
.. _setup:
 

	
 
Setup
 
=====
 

	
 

	
 
Setting up the application
 
--------------------------
 

	
 
::
 
 
 
 paster make-config RhodeCode production.ini
 

	
 
- This will create `production.ini` config inside the directory
 
  this config contains various settings for RhodeCode, e.g proxy port, 
 
  email settings,static files, cache and logging.
 

	
 
::
 

	
 
 paster setup-app production.ini
 

	
 
- This command will create all needed tables and an admin account. 
 
  When asked for a path You can either use a new location of one with already 
 
  existing ones. RhodeCode will simply add all new found repositories to 
 
  it's database. Also make sure You specify correct path to repositories.
 
- Remember that the given path for mercurial_ repositories must be write 
 
  accessible for the application. It's very important since RhodeCode web interface
 
  will work even without such an access but, when trying to do a push it'll 
 
  eventually fail with permission denied errors. 
 
- Run 
 

	
 
::
 
 
 
 paster serve production.ini
 
 
 
- This command runs the RhodeCode server the app should be available at the 
 
  127.0.0.1:5000. This ip and port is configurable via the production.ini 
 
  file  created in previous step
 
- Use admin account you created to login.
 
- Default permissions on each repository is read, and owner is admin. So 
 
  remember to update these if needed.
 
  
 
    
 
Setting up Whoosh full text search
 
----------------------------------
 

	
 
Index for whoosh can be build starting from version 1.1 using paster command
 
passing repo locations to index, as well as Your config file that stores
 
whoosh index files locations. There is possible to pass `-f` to the options
 
to enable full index rebuild. Without that indexing will run always in in
 
incremental mode.
 

	
 
::
 

	
 
 paster make-index --repo-location=<location for repos> production.ini  
 

	
 
for full index rebuild You can use
 

	
 
::
 

	
 
 paster make-index -f --repo-location=<location for repos> production.ini
 

	
 
- For full text search You can either put crontab entry for
 

	
 
This command can be run even from crontab in order to do periodical 
 
index builds and keep Your index always up to date. An example entry might 
 
look like this
 

	
 
::
 
 
 
 /path/to/python/bin/paster --repo-location=<location for repos> /path/to/rhodecode/production.ini
 
  
 
When using incremental(default) mode whoosh will check last modification date 
 
of each file and add it to reindex if newer file is available. Also indexing 
 
daemon checks for removed files and removes them from index. 
 

	
 
Sometime You might want to rebuild index from scratch. You can do that using 
 
the `-f` flag passed to paster command or, in admin panel You can check 
 
`build from scratch` flag.
 

	
 

	
 
Setting up LDAP support
 
-----------------------
 

	
 

	
 
RhodeCode starting from version 1.1 supports ldap authentication. In order
 
to use ldap, You have to install python-ldap package. This package is available
 
via pypi, so You can install it by running
 

	
 
::
 

	
 
 easy_install python-ldap
 
 
 
::
 

	
 
 pip install python-ldap
 

	
 

	
 
python-ldap requires some certain libs on Your system, so before installing it 
 
check that You have at least `openldap`, and `sasl` libraries.
 

	
 
ldap settings are located in admin->permissions section,
 

	
 
Here's a typical ldap setup::
 

	
 
 Enable ldap  = checked                 #controlls if ldap access is enabled
 
 Host         = host.domain.org         #acctuall ldap server to connect
 
 Port         = 389 or 689 for ldaps    #ldap server ports
 
 Enable LDAPS = unchecked               #enable disable ldaps
 
 Account      = <account>               #access for ldap server(if required)
 
 Password     = <password>              #password for ldap server(if required)
 
 Base DN      = CN=users,DC=host,DC=domain,DC=org
 
 
 

	
 
`Account` and `Password` are optional, and used for two-phase ldap 
 
authentication so those are credentials to access Your ldap, if it doesn't 
 
support anonymous search/user lookups.
 

	
 
If all data are entered correctly, and `python-ldap` is properly installed
 
Users should be granted to access RhodeCode wit theire ldap accounts. When 
 
logging at the first time an special ldap account is created inside RhodeCode, 
 
so You can control over permissions even on ldap users. If such user exists 
 
already in RhodeCode database ldap user with the same username would be not 
 
able to access RhodeCode.
 

	
 
If You have problems with ldap access and believe You entered correct 
 
information check out the RhodeCode logs,any error messages sent from 
 
ldap will be saved there.
 

	
 

	
 
Nginx virtual host example
 
--------------------------
 

	
 
Sample config for nginx using proxy::
 

	
 
 server {
 
    listen          80;
 
    server_name     hg.myserver.com;
 
    access_log      /var/log/nginx/rhodecode.access.log;
 
    error_log       /var/log/nginx/rhodecode.error.log;
 
    location / {
 
            root /var/www/rhodecode/rhodecode/public/;
 
            if (!-f $request_filename){
 
                proxy_pass      http://127.0.0.1:5000;
 
            }
 
            #this is important for https !!!
 
            proxy_set_header X-Url-Scheme $scheme;
 
            include         /etc/nginx/proxy.conf;  
 
    }
 
 }  
 
  
 
Here's the proxy.conf. It's tuned so it'll not timeout on long
 
pushes and also on large pushes::
 

	
 
    proxy_redirect              off;
 
    proxy_set_header            Host $host;
 
    proxy_set_header            X-Host $http_host;
 
    proxy_set_header            X-Real-IP $remote_addr;
 
    proxy_set_header            X-Forwarded-For $proxy_add_x_forwarded_for;
 
    proxy_set_header            Proxy-host $proxy_host;
 
    client_max_body_size        400m;
 
    client_body_buffer_size     128k;
 
    proxy_buffering             off;
 
    proxy_connect_timeout       3600;
 
    proxy_send_timeout          3600;
 
    proxy_read_timeout          3600;
 
    proxy_buffer_size           8k;
 
    proxy_buffers               8 32k;
 
    proxy_busy_buffers_size     64k;
 
    proxy_temp_file_write_size  64k;
 
 
 
Also when using root path with nginx You might set the static files to false
 
in production.ini file::
 

	
 
  [app:main]
 
    use = egg:rhodecode
 
    full_stack = true
 
    static_files = false
 
    lang=en
 
    cache_dir = %(here)s/data
 

	
 
To not have the statics served by the application. And improve speed.
 

	
 
Apache reverse proxy
 
--------------------
 
Tutorial can be found here
 
http://wiki.pylonshq.com/display/pylonscookbook/Apache+as+a+reverse+proxy+for+Pylons
 

	
 

	
 
Apache's example FCGI config
 
----------------------------
 

	
 
TODO !
 

	
 
Other configuration files
 
-------------------------
 

	
 
Some extra configuration files and examples can be found here:
 
http://hg.python-works.com/rhodecode/files/tip/init.d
 

	
 
and also an celeryconfig file can be use from here:
 
http://hg.python-works.com/rhodecode/files/tip/celeryconfig.py
 

	
 
Troubleshooting
 
---------------
 

	
 
- missing static files ?
 

	
 
 - make sure either to set the `static_files = true` in the .ini file or
 
   double check the root path for Your http setup. It should point to 
 
   for example:
 
   /home/my-virtual-python/lib/python2.6/site-packages/rhodecode/public
 
   
 
- can't install celery/rabbitmq
 

	
 
 - don't worry RhodeCode works without them too. No extra setup required
 

	
 

	
 
- long lasting push timeouts ?
 

	
 
 - make sure You set a longer timeouts in Your proxy/fcgi settings, timeouts
 
   are caused by https server and not RhodeCode
 

	
 
- large pushes timeouts ?
 
 
 
 - make sure You set a proper max_body_size for the http server
 

	
 

	
 

	
 
.. _virtualenv: http://pypi.python.org/pypi/virtualenv
 
.. _python: http://www.python.org/
 
.. _mercurial: http://mercurial.selenic.com/
 
.. _celery: http://celeryproject.org/
 
.. _rabbitmq: http://www.rabbitmq.com/
 
\ No newline at end of file
0 comments (0 inline, 0 general)