Incidents/2025-07-21 Cirrussearch outage
document status: draft
Summary
| Incident ID | 2025-07-21 Cirrussearch outage | Start | 2025-07-21 23:51 |
|---|---|---|---|
| Task | T400160 | End | 2025-07-22 00:17 |
| People paged | 0 | Responder count | 2 |
| Coordinators | None | Affected metrics/SLOs | |
| Impact | From 22025-07-21 23:51 to 2025-07-21 23:57, any on-wiki searches pointed at the EQIAD datacenter (which typically serves the majority of production traffic) likely failed. | ||
Timeline
The following is taken from the SAL and #wikimedia-search IRC.
14:28 inflatador (Brian King, Data Platform SRE and owner of the Cirrussearch service) starts a rolling restart of the EQIAD search cluster to apply some config changes .
17:03 The rolling restart script restarts cirrussearch1100. Cirrussearch1100 is a master-eligible host, one of five responsible for maintaining cluster state in our production EQIAD cluster. The host comes back up with outdated master data, which causes split-brain in the cluster. All API calls to the cluster begin to fail.
17:06 inflatador depools the EQIAD datacenter, ending user impact.
17:10 ebernhardson (Search Platform SWE) discovers log entries pointing to outdated master eligible data on cirrussearch1100.
17:19 inflatador stops OpenSearch on cirrussearch1100, which stops it from trying to join the cluster with bad data. The cluster regains quorum and starts to recover.
17:36 inflatador removes the datadir on cirrussearch1100 and starts the node; it joins the cluster cleanly this time.
17:40 inflatador repools the EQIAD cluster, returning the search services to full capacity.
Did the appropriate alert(s) fire? Was the alert volume manageable? Did they point to the problem with as much accuracy as possible?
Detection
How was the issue was first detected? By me (inflatador).
Did the appropriate alert(s) fire? Yes
Was the alert volume manageable? Yes
Did they point to the problem with as much accuracy as possible? Not exactly, we got timeout alerts and we probably want to have verbiage that mentions the exact error from OpenSearch (503/no cluster leader).
Conclusions
What went well?
- Failover happened quickly
What went poorly?
- This probably happened because we did not follow the procedure for replacing master-eligibles during the recent OpenSearch migration.
- The fix for this issue was not documented on our administration page
Where did we get lucky?
Links to relevant documentation
- …
Add links to information that someone responding to this alert should have (runbook, plus supporting docs). If that documentation does not exist, add an action item to create it.
Actionables
- …
Create a list of action items that will help prevent this from happening again as much as possible. Link to or create a Phabricator task for every step.
Add the #Sustainability (Incident Followup) and the #SRE-OnFire Phabricator tag to these tasks.
Scorecard
| Question | Answer
(yes/no) |
Notes | |
|---|---|---|---|
| People | Were the people responding to this incident sufficiently different than the previous five incidents? | ||
| Were the people who responded prepared enough to respond effectively | |||
| Were fewer than five people paged? | |||
| Were pages routed to the correct sub-team(s)? | |||
| Were pages routed to online (business hours) engineers? Answer “no” if engineers were paged after business hours. | |||
| Process | Was the "Incident status" section atop the Google Doc kept up-to-date during the incident? | ||
| Was a public wikimediastatus.net entry created? | |||
| Is there a phabricator task for the incident? | |||
| Are the documented action items assigned? | |||
| Is this incident sufficiently different from earlier incidents so as not to be a repeat occurrence? | |||
| Tooling | To the best of your knowledge was the open task queue free of any tasks that would have prevented this incident? Answer “no” if there are open tasks that would prevent this incident or make mitigation easier if implemented. | ||
| Were the people responding able to communicate effectively during the incident with the existing tooling? | |||
| Did existing monitoring notify the initial responders? | |||
| Were the engineering tools that were to be used during the incident, available and in service? | |||
| Were the steps taken to mitigate guided by an existing runbook? | |||
| Total score (count of all “yes” answers above) | |||