Backup and recovery testing
Know what it takes to restore your application
Have backups, but no recent evidence that your application can use them? Start with one defined workload: review the backup path, rehearse recovery in an isolated environment, and receive a tested runbook with measured results and remaining gaps.
Start with a short brief. A call is optional. Project scope, price and schedule are agreed before work starts.
What your project delivers
Choose the work your environment needs. The proposal records which of these artifacts are included.
A workload and recovery map
List the database, files, configuration, identities, keys and external dependencies the application needs. Agree the failure scenario, acceptable data loss (RPO) and acceptable service interruption (RTO) with the owner.
A backup plan tied to the workload
Review the backup method, retention, encryption, access and recovery location. Record missing dependencies and any agreed configuration changes. A replica or a successful scheduled job alone is not a restore test.
An isolated restore rehearsal
Restore an agreed recovery point into a separate test environment. Check data and representative application paths, prevent test systems from sending real customer notifications, and record what failed as well as what passed.
Evidence and an operator runbook
Receive the recovery-point timestamp, measured restore and validation durations, test results, blockers and repeatable instructions. Name the owner of recurring tests, backup alerts and any remediation work.
What we need to get started
- Describe one application, its database engine/version, hosting provider, data volume and current backup method. AWS-managed and self-hosted workloads need different recovery procedures; discovery confirms fit.
- Tell us when a restore was last tested, what changed since then and which business workflow must work after recovery. If the answer is unknown, say so in the brief.
- Name the application owner and agree recovery objectives, permitted data locations, test-data handling, isolation and a budget for temporary infrastructure.
- Start with a description of the setup. Any access, restore operation or infrastructure change follows your approval process; the enquiry needs no credentials or backup files.
Use the initial enquiry to describe the problem. Keep passwords, access keys and customer records out of the form. Any required access is arranged separately through your approved process.
From discovery to handover
Step 1
Agree one testable outcome
Choose the workload and failure scenario. Record the scope, access, recovery objectives, price, test window and stop conditions before work begins.
Step 2
Prepare the recovery path
Check backup availability, keys, permissions, compatible software and target capacity. Review isolation and cleanup before restoring data.
Step 3
Restore and validate
Run the agreed rehearsal, measure the steps and exercise representative application checks. Separate infrastructure restoration from application readiness and end-to-end service recovery.
Step 4
Hand over the evidence
Walk through the runbook with the receiving operator. Record passed checks, failed objectives, changes needed, costs and the next test owner. Remove temporary access and confirm agreed test-resource cleanup.
How we agree that the work is done
- The chosen recovery point can be restored into the agreed isolated environment, or the exact blockers and corrective work are documented. A failed test is not reported as successful recovery.
- Agreed data-integrity and application checks pass with results recorded. Database startup by itself does not satisfy application recovery.
- Measured data freshness and recovery durations are compared with the agreed RPO and RTO, using explicit start and end points. A rehearsal result is not a guaranteed production recovery time.
- The owner receives the runbook, dependencies, access requirements, evidence, remaining risks and repeat-test schedule. Temporary resources and access are accounted for.
Plan your recovery rehearsal
Use the blank worksheet to describe your stack, recovery objectives and the evidence you have today. Mark anything you have not tested as unknown so we can scope the missing checks. Download it without an email signup.
Download the backup and restore project worksheet
Which backup, what timestamp, and which keys and permissions are required?
Which data checks and user workflow demonstrate that the restored application works?
What passed, what remains unresolved, and who runs the next rehearsal?
For PostgreSQL, include database roles and other dependencies in the plan: a single-database dump does not include cluster-wide roles. The PostgreSQL backup documentation explains these boundaries. For supported AWS resources, AWS Backup restore testing can automate restore jobs; application validation still needs an explicit plan.
For the architecture behind failover, read our PostgreSQL HA guide. High availability and restoring a historical backup answer different failure scenarios.
Scope, costs and ongoing ownership
This is planned engineering work for an agreed workload, not an emergency recovery hotline. Scope and price depend on data volume, backup condition, dependencies and isolation needs. Test infrastructure, storage, transfer and software costs are separate. Recovering unavailable or corrupted data, a full multi-region redesign, compliance certification and staffed incident response are not included unless separately agreed. Ongoing operation needs its own support arrangement.
A project has agreed milestones and a handover. If you need continuing operation, compare our managed DevOps service and monthly retainer. Ongoing support and staffed coverage are agreed separately.
Explore the technical decisions
Bring the problem you want to solve
Tell us about your stack, the delivery or reliability problem, the environments involved and your target date. We review the brief to establish fit, identify missing inputs and agree the next scoping step. A detailed review or implementation needs its own agreed scope.
Scope a backup and recovery projectPrefer email? hello@instadevops.com