Page MenuHomePhabricator

Improve FrontendUnavailable alerts with more information/context of what's failing
Closed, ResolvedPublic

Description

During the incident on 2023-02-22 two FrontendUnavailable alerts fired, however it wasn't clear exactly what was failing (i.e. varnish and ats were, globally)

  • FrontendUnavailable (varnish-text)
  • FrontendUnavailable cache_text ()

Event Timeline

Change 892362 had a related patch set uploaded (by Filippo Giunchedi; author: Filippo Giunchedi):

[operations/alerts@master] sre: more readable varnish/haproxy frontend unavailable

https://gerrit.wikimedia.org/r/892362

Change 892362 merged by Filippo Giunchedi:

[operations/alerts@master] sre: more readable varnish/haproxy frontend unavailable

https://gerrit.wikimedia.org/r/892362

fgiunchedi claimed this task.

This is done, we now have more descriptive alerts based on what's failing (haproxy/varnish)