Page MenuHomePhabricator

Create an alert for depooled cp hosts
Open, Needs TriagePublic

Description

Currently, if a host is depooled with confctl, it is possible for that host to be left in a depooled state indefinitely. This task is for the creation of a Prometheus alert that would fire if a single host is left depooled for longer than some period of time (to be determined).
Note: this purposely does not cover or account for cookbook-depooled hosts being left in that state

Event Timeline

Change #1219634 had a related patch set uploaded (by CDobbins; author: CDobbins):

[operations/puppet@production] prometheus: add depooled cp* host check

https://gerrit.wikimedia.org/r/1219634

Change #1219634 merged by CDobbins:

[operations/puppet@production] prometheus: add pooled host check

https://gerrit.wikimedia.org/r/1219634

Change #1249385 had a related patch set uploaded (by CDobbins; author: CDobbins):

[operations/puppet@production] prometheus: fix pooled host check

https://gerrit.wikimedia.org/r/1249385

Change #1249385 merged by CDobbins:

[operations/puppet@production] prometheus: fix pooled host check

https://gerrit.wikimedia.org/r/1249385