Control plane¶
- PG-15 Control plane down: what runs?: A control-plane outage is not a workload outage. Protect what runs, then repair.
- PG-16 MessagingTimeout everywhere?: Computes reported down are a symptom here. Read every RabbitMQ node before you reset anything.
- PG-17 APIs fail on database errors?: The data is at risk only during the recovery. Know the most advanced node before you bootstrap.
- PG-26 The API answers 503?: A 503 from HAProxy is a messenger. Find out why the backends fail their check.
- PG-27 401 with a valid token?: A fresh token hides the problem. Test with one older than the token cache.
- PG-28 Out of database connections?: Three limits, three messages. The message tells you where to look first.
- PG-42 Conductor or scheduler down?: With the conductor gone, every compute looks down. Nothing is wrong on the hosts.
- PG-43 Provisioning fails everywhere?: When everything fails at once, look at what everything shares.