kallithea Files · docs/usage/debugging.rst

Files @ 6c3bda995a88

Branch filter:

Location: kallithea/docs/usage/debugging.rst

6c3bda995a88 1.2 KiB text/prs.fallenstein.rst Show Annotation Show as Raw Download as Raw

domruf

js: use ajax requests for select2 autocomplete

When you have a big user base, with thousends of users, always using the whole
dataset makes the UI slow.

This will replace kallithea/model/repo.py get_users_js and get_user_groups_js
which were used to inline the full list of users and groups in the document.
Instead, it will expose a json service for doing the completion.

When using the autocomplete, there might be multiple ajax requests, but tests
with a userbase > 9000 showed no problems.
And keep in mind, that although we now make multiple requests (one for every
character) that
- the autocomplete is not used that often
- the requests are quite cheap
- most importanly, we no longer need to calculate the user list/group list if
the user doesn't use the autocomplete

Users and groups are still passed as parameters to the javascript functions -
they will be removed later.

.. _debugging:

===================
Debugging Kallithea
===================

If you encounter problems with Kallithea, here are some instructions
on how to debug them.

.. note:: First make sure you're using the latest version available.


Enable detailed debug
---------------------

Kallithea uses the standard Python ``logging`` module to log its output.
By default only loggers with ``INFO`` level are displayed. To enable full output
change ``level = DEBUG`` for all logging handlers in the currently used .ini file.
This change will allow you to see much more detailed output in the log file or
console. This generally helps a lot to track issues.


Enable interactive debug mode
-----------------------------

To enable interactive debug mode simply comment out ``set debug = false`` in
the .ini file. This will trigger an interactive debugger each time
there is an error in the browser, or send a http link if an error occurred in the backend. This
is a great tool for fast debugging as you get a handy Python console right
in the web view.

.. warning:: NEVER ENABLE THIS ON PRODUCTION! The interactive console
             can be a serious security threat to your system.