Operations

Use this section once HestiaStore is already integrated and you need to run it reliably in production-like environments.

What belongs here

• WAL enablement and recovery
• monitoring and alerting
• tuning for memory, I/O, and maintenance behavior
• backup, restore, and post-crash validation
• staged rollout and rollback procedures

Recommended operating path

1. Start with WAL if you need local crash recovery.
2. Add Monitoring before broad rollout.
3. Use Performance Tuning only after you have workload evidence.
4. Define Backup & Restore before calling the system production-ready.
5. Use WAL Canary Runbook for staged WAL rollouts.

After an unexpected shutdown

Recommended recovery sequence:

1. Reopen the index.
2. Run checkAndRepairConsistency().
3. If WAL is enabled, inspect the WAL directory with wal_verify.
4. Run compact() if you need to restore locality and clean up fragmented layout.
5. Take a fresh backup after the system is healthy again.

Related docs

• Configuration for builder-level settings
• Monitoring Architecture for internal monitoring design
• Performance Model & Sizing for implementation-level tuning context