Skip to content

Incidents

Post-incident write-ups for notable failures on the kensan-lab cluster — what broke, the timeline, the root cause, and the follow-up actions. They exist so the same failure mode is recognized faster next time.

Why so short

Formal write-ups here start from 2026-03. Earlier failure modes (namespace prune, PV loss under Retain-less reclaim policies, ArgoCD ownership takeover, etc.) are captured as rationale inside the relevant ADRs rather than as standalone incidents — that backlog is being migrated over time.

Index

Date Incident
2026-03-02 system-infra Sync Blocked