CI workloads impacted by GCP maintenance event (Severity 2)

Problem

CI workloads are being delayed by a GCP maintenance event.

Impact

CI jobs were delayed, with runner queue times increasing and pending jobs rising into the thousands. This also affected deployments and feature flag changes.

Causes

GCP maintenance in an affected zone reduced runner capacity.

Response strategy

We merged a change to avoid the maintenance-affected zone and rolled the affected pods. Queue delays and pending jobs have started to recover.

Post-incident review

Incident review, owned by @kkyrala. FCL exception request; decision pending.