DexaflowSchemaNotReady
What DexaflowSchemaNotReady means and what to do when it fires.
Symptom. A control plane pod stays up and not ready because the database schema is not one its binary can serve: absent, behind, dirty, unreadable, or the database was replaced (ADR 0090).
Impact. The pod takes no traffic and, if it would lead, schedules nothing.
database replaced is critical: the control plane points at a database that is
not the one it ran on.
First commands.
kubectl -n <namespace> get pods -l app.kubernetes.io/name=dexaflow
kubectl -n <namespace> logs deploy/<release> --tail=20 | grep 'database schema'
kubectl -n <namespace> logs -l app.kubernetes.io/component=migrate --tail=50
Decide.
schema absentorschema behind: the migration Job did not run or failed. Read its log and rerun the upgrade.schema dirty: a migration stopped halfway. Follow The migration Job.database replaced: follow Database lost, restored or repointed.