DATABASES.HOW — OPERATIONS CHECKLIST
0 of 10 checks completed
Self-assessment, not a certification or production approval.

[ ] Agree recovery targets
    Write down acceptable data loss (RPO) and time to recover (RTO), and get the owner’s agreement.

[ ] Protect independent backups
    Automate encrypted backups, monitor failures, and keep a recoverable copy outside the primary failure boundary.

[ ] Complete a restore drill
    Restore to an isolated target, verify application data, and record the measured recovery time.

[ ] Monitor the workload
    Track latency, errors, connections, replication lag, slow queries, and resource saturation.

[ ] Test capacity and failure
    Exercise representative peak traffic and node failure with room for maintenance and growth.

[ ] Review access and protection
    Use least privilege, encrypted connections, tested credential rotation, and restricted network access.

[ ] Rehearse changes
    Test migrations, compatibility, rollback options, and maintenance windows on representative data.

[ ] Name an operator
    Assign an owner, on-call coverage, escalation contacts, and accessible runbooks.

[ ] Set a complete budget
    Include compute, storage, backups, transfer, licenses, and people; enable spend alerts.

[ ] Document the data lifecycle
    Define retention, deletion, exports, and how backup retention interacts with deletion requirements.

Owner: ___
Date of review: ___
Evidence / next action for open items: ___

Restore guide: https://databases.how/guides/test-your-database-restore