-
Type:
Task
-
Resolution: Won't Do
-
Priority:
Major - P3
-
None
-
Affects Version/s: None
-
Component/s: None
-
None
-
Storage Execution
-
Storage Execution 2026-11-23
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Add structured logs that make it possible to determine when each standby/non-primary is actively performing internode hash validation and when validation coverage is unavailable, particularly during startup, recovery, checkpoint adoption, and other non-steady-state periods.
Motivation
A standby can crash after adopting checkpoint N and later restart from checkpoint N+10. Adopting the newer checkpoint does not itself demonstrate that hashes were validated over the intervening window. We need visibility into this gap and when validation resumes.
Scope
- Log validation state at startup and whenever it changes, including entering or leaving active validation, with a reason for the transition.
- Include node identity, role, wall-clock event time, adopted checkpoint identity/valid-as-of timestamp, and the last successfully validated checkpoint/valid-as-of timestamp, when known.
- Log successful checkpoint validations with the logical interval covered and the wall-clock completion time. Distinguish adopting a baseline from validating an interval.
- Identify known uncovered logical intervals when checkpoint adoption or recovery skips validation. Explicitly report unknown prior coverage after restart rather than implying continuous validation.
- Distinguish full, partial, and unavailable hash coverage where collections are excluded or hashes are missing.
- Record when validation is disabled for repair and when protection actually resumes. We may want to put timestamps for when log-only mode is also in effect.
- another helpful thing to log is when we establish baseline for the standby hash.