2025-09-24: Apdex SLO violation for SQL transactions from Rails to Postgres primary

Apdex SLO violation for SQL transactions from Rails to Postgres primary (Severity 3 (Medium))

Problem: The Apdex score for SQL transactions from the Rails monolith to the Postgres primary fell below the SLO, following a spike in database load and transaction time.

Impact: SQL transaction performance between the Rails application and the Postgres primary dropped below the SLO threshold, with Apdex falling to 98.28%. This affected the main production database cluster and may have degraded user experience during the incident window.

Causes: A spike in database load and transaction time on a replica node was observed between 12:50 and 12:58. Shortly after, similar degraded performance appeared on the primary node. The team is investigating if a VM migration or another cluster operation triggered these symptoms.

Response strategy: An issue has been raised to cease creation of FKs on partitions of p_duo_workflows_checkpoints gitlab-org/gitlab#572109 (closed)


This ticket was created to track INC-4152, by incident.io 🔥