2025-09-25: Sidekiq queueing SLO violation on low-urgency-cpu-bound shard

Sidekiq queueing SLO violation on low-urgency-cpu-bound shard (Severity 4 (Low))

Problem: Sidekiq jobs on the low-urgency-cpu-bound shard do not meet their queueing time SLO.

Impact: No customer impact has been observed. However, repeated alerts from this SLO violation have increased the workload for the SRE on-call team.

Causes: A feature flag for Sidekiq concurrency limits on the low-urgency-cpu-bound shard was set too low. This caused PipelineHooksWorker jobs to be deferred, resulting in a drop in the proportion of jobs meeting queueing time targets.

Response strategy: We are reverting the relevant feature flags as a controlled rollback to improve queueing performance and restore the Apdex. We are also scaling up the shard to improve the queueing apdex as we’ve seen the shard is saturated on the available workers sidekiq_shard_workers.


This ticket was created to track INC-4208, by incident.io 🔥