Tech:Troubleshooting: Difference between revisions

From Orain Meta
Jump to navigation Jump to search
Content added Content deleted
(Error in the word erorrs xd)
(→‎Ansible Playbook doesn't appear to be updating: error in errors in erorrs what is this :/)
Line 14: Line 14:
===Ansible Playbook doesn't appear to be updating===
===Ansible Playbook doesn't appear to be updating===
# Is the cronjob for root still there?
# Is the cronjob for root still there?
# Check /var/log/ansible-pull.log for error
# Check /var/log/ansible-pull.log for errors
# Fix them (best bet is probably reverting the changes on github that broke it!)
# Fix them (best bet is probably reverting the changes on github that broke it!)



Revision as of 15:46, 28 May 2014

UPDATE THE SAL WITH WHAT YOU DO!

This is a cached copy of the requested page, and may not be up to date (in red)

  1. Try running 'rebuildFileCache.php' over the wiki
  2. If you get the message saying the wiki does not exist see Renaming a database
  3. If the issue persists, fix it!

Mediawiki Magic Word Exceptions

This probably means that the Localisation cache has not updated or someone has enabled an extension on a wiki without it being enabled on extload.

  1. Make sure the extension is included on extloadwiki
  2. If the issue persists run 'rebuildLocalisationCache.php --wiki extloadwiki'
  3. If the issue persists check /var/log/mediawiki/exception.log and fix it!

Ansible Playbook doesn't appear to be updating

  1. Is the cronjob for root still there?
  2. Check /var/log/ansible-pull.log for errors
  3. Fix them (best bet is probably reverting the changes on github that broke it!)

You changed stuff, stuff broke, you dont know why

  1. If it was an extension have you read the documentation and done everything required (update.php, dependancies)
  2. Have you done everything documented currently at our readme
  3. Try manually flushing the APC cache see here
  4. Revert your change! or cry and scream until someone else fixes it

Server/nginx dies or is unresponsive

When this happened the web server seemed to totally inaccessible. Hitting any page would not even show / throw a MW exception, nothing would appear in the nginx, mediawiki or syslogs...

  1. nginx stop
  2. /etc/init.d/nginx start
  3. rebuild i18n if it works - if not, looks as the nginx error.log or syslogs.