Skip to content
This repository has been archived by the owner on Nov 30, 2022. It is now read-only.

Commit

Permalink
Fix some _crazy_ text in session.rst
Browse files Browse the repository at this point in the history
Not sure how that even got that way!  Also updated the docs script to
create the temp build dir if not there already.
  • Loading branch information
jmcfarlane committed Jan 10, 2010
1 parent defbe4c commit c9eef3f
Show file tree
Hide file tree
Showing 3 changed files with 28 additions and 27 deletions.
4 changes: 2 additions & 2 deletions docs/about.rst
Original file line number Diff line number Diff line change
Expand Up @@ -75,8 +75,8 @@ find defects or have ideas for improvement please feel free to file
issues. In the event Chula becomes more popular, a more sophisticated
tracker will be used.

The link to the Github tracker is `here
<http://github.com/jmcfarlane/chula/issues>`_.
The link to the Github tracker is here:
http://github.com/jmcfarlane/chula/issues

Release cycle
+++++++++++++
Expand Down
46 changes: 25 additions & 21 deletions docs/session.rst
Original file line number Diff line number Diff line change
Expand Up @@ -23,17 +23,19 @@ have to use `sticky sessions` in your load balancer configuration.
Scalable
++++++++

10% of the traffic. Because session is usually transient in nature,
all requests to the cache, and only send requests to the backend Chula
session is reasonably scalable. The default behavior is to send
configurations Memcached is used as the cache. Memcached has a
dramatically. The default configuration is to send the backend only
periodically. This means that you can reduce your backend load
reputation as performing well. this is a reasonably safe way to
increase performance. The current limitation of the session store is
that it does not use connection pooling. This can optionally be added
by fronting your PostgreSQL server with pgpool_. When CouchDB is used
as the backend, connection pooling isn't relevent as it uses HTTP.
Chula maintains session in Memcached_ backed by a persistent data
store. Because Memcache is reasonably reliable, only a percentage of
session requests are actually sent to the backend. The default
configuration is to update the backend every 10 requests. This means
the session backend (which is slower than cache) is only serving 10%
of the actual traffic. The frequency of backend updates is
configurable via :attr:`config.Config.session_max_stale_count`, and
the backend is always consulted in the event of a cache miss.

A current limitation of the session store is that it does not use
connection pooling. This can optionally be added by fronting your
PostgreSQL server with pgpool_. When CouchDB is used as the backend,
connection pooling isn't relevent as it uses HTTP.

Native Storage
++++++++++++++
Expand Down Expand Up @@ -73,31 +75,33 @@ Assuming you're connecting to the server running locally, you're all
done!

If you're connecting to a remote server, you need to add the hostname
in your configuration via ``session_host``. If you want to use a different user/password/port/server you are free
to do so.
in your configuration via :attr:`config.Config.session_host`. If you
want to use a different user/password/port/server you are free to do
so.

CouchDB
-------

You wil need to configure ``session_nosql`` with the full HTTP path to
your CouchDB server. If this is a local install, you'd set the value
to http://localhost:5984. Don't worry about the database, as it wil
be created automatically on demand.
You wil need to configure :attr:`config.Config.session_nosql` with the
full HTTP path to your CouchDB server. If this is a local install,
you'd set the value to http://localhost:5984. Don't worry about the
database, as it wil be created automatically on demand.

Memcached
---------

You need to configure your cluster information via
``session_memcache``. If you're using a local install of Memcached
you can just take the defaults. Else, specify it with something like
this::
:attr:`config.Config.session_memcache`. If you're using a local
install of Memcached you can just take the defaults, else configure it
with something like this::

[('host1:11211', 1), ('host2:11211', 1), ('host3:11211', 1)]

Reference
+++++++++

For more detail on Chula configuration in general, see here_.
For more detail on Chula configuration in general, see
:mod:`config`.

.. _here: configuration.html
.. _pgpool: http://pgpool.projects.postgresql.org/
Expand Down
5 changes: 1 addition & 4 deletions scripts/docs
Original file line number Diff line number Diff line change
Expand Up @@ -6,10 +6,7 @@ SRC=$(dirname $0)
cd $SRC/..

# Validate the base dir exists
if [ ! -d "$BASEDIR" ]; then
echo "$BASEDIR does not exist, fix and try again"
exit 1
fi
mkdir -p "$BASEDIR"

# Fetch the current version for which docs are being generated
version=$(grep -o -E '[0-9]+\.[0-9]+\.[0-9]+[a-zA-Z_-]*' chula/__init__.py)
Expand Down

0 comments on commit c9eef3f

Please sign in to comment.