Files @ b537babcf966
Branch filter:

Location: kallithea/docs/usage/performance.rst - annotation

b537babcf966 2.5 KiB text/prs.fallenstein.rst Show Source Show as Raw Download as Raw
Søren Løvborg
login: include query parameters in came_from

The login controller uses the came_from query argument to determine
the page to continue to after login.

Previously, came_from specified only the URL path (obtained using
h.url.current), and any URL query parameters were passed along as
separate (additional) URL query parameters; to obtain the final redirect
target, h.url was used to combine came_from with the request.GET.

As of this changeset, came_from specifies both the URL path and query
string (obtained using request.path_qs), which means that came_from can
be used directly as the redirect target (as always, WebOb handles the
task of expanding the server relative path to a fully qualified URL).
The mangling of request.GET can also be removed.

The login code appended arbitrary, user-supplied query parameters to
URLs by calling the Routes URLGenerator (h.url) with user-supplied
keyword arguments. This construct is unfortunate, since url only
appends _unknown_ keyword arguments as query parameters, and the
parameter names could overlap with known keyword arguments, possibly
affecting the generated URL in various ways. This changeset removes
this usage from the login code, but other instances remain.

(In practice, the damage is apparently limited to causing an Internal
Server Error when going to e.g. "/_admin/login?host=foo", since WebOb
returns Unicode strings and URLGenerator only allows byte strings for
these keyword arguments.)
.. _performance:

================================
Optimizing Kallithea performance
================================

When serving a large amount of big repositories, Kallithea can start
performing slower than expected. Because of the demanding nature of handling large
amounts of data from version control systems, here are some tips on how to get
the best performance.

* Kallithea is often I/O bound, and hence a fast disk (SSD/SAN) is
  usually more important than a fast CPU.

* Sluggish loading of the front page can easily be fixed by grouping repositories or by
  increasing cache size (see below). This includes using the lightweight dashboard
  option and ``vcs_full_cache`` setting in .ini file.

Follow these few steps to improve performance of Kallithea system.

1. Increase cache

    Tweak beaker cache settings in the ini file. The actual effect of that
    is questionable.

2. Switch from SQLite to PostgreSQL or MySQL

    SQLite is a good option when having a small load on the system. But due to
    locking issues with SQLite, it is not recommended to use it for larger
    deployments. Switching to MySQL or PostgreSQL will result in an immediate
    performance increase. A tool like SQLAlchemyGrate_ can be used for
    migrating to another database platform.

3. Scale Kallithea horizontally

    Scaling horizontally can give huge performance benefits when dealing with
    large amounts of traffic (many users, CI servers, etc.). Kallithea can be
    scaled horizontally on one (recommended) or multiple machines. In order
    to scale horizontally you need to do the following:

    - Each instance needs its own .ini file and unique ``instance_id`` set.
    - Each instance's ``data`` storage needs to be configured to be stored on a
      shared disk storage, preferably together with repositories. This ``data``
      dir contains template caches, sessions, whoosh index and is used for
      task locking (so it is safe across multiple instances). Set the
      ``cache_dir``, ``index_dir``, ``beaker.cache.data_dir``, ``beaker.cache.lock_dir``
      variables in each .ini file to a shared location across Kallithea instances
    - If celery is used each instance should run a separate Celery instance, but
      the message broker should be common to all of them (e.g.,  one
      shared RabbitMQ server)
    - Load balance using round robin or IP hash, recommended is writing LB rules
      that will separate regular user traffic from automated processes like CI
      servers or build bots.


.. _SQLAlchemyGrate: https://github.com/shazow/sqlalchemygrate