Isolated restore test

Recovery Proof

Test one backup chain on an isolated target. Get a written record of what happened.

For teams with a backup chain, a suitable isolated SQL Server target and an approved test window.

Get a record of the restore attempt, its timings and its success or failure. The report explains which checks ran and the limits of what they demonstrate.

See scope & prices Check which service fits

One-time service · Scope confirmed before payment

Dark mechanical keyboard with orange accent keys on a technical workspace.

Source database: 1 · up to 100 GB allocated data
Isolated target: 1 customer-supplied instance
Test window: 3 hours active · 4 hours elapsed
Engagement: One scheduled test

How it works

Managed only. An isolated restore test.

Find out whether the selected backups restore on the agreed target, which checks pass and what needs attention next.

An agreed test in your environment.

Managed

Your team supplies the selected backup chain, an isolated SQL Server target and approved access. We carry out the agreed restore attempt and checks in the booked window. Before payment, we must check the chain, target and test requirements and confirm the exact scope and fixed quote.

Standard: From $1,695
Priority Same scope, reserved scheduling: From $2,445

USD · excluding applicable taxes · per engagement · confirmed quote required

You receive up to three hours of active technical work and four hours elapsed in the test window, one restore attempt, the procedural correction and retry when it qualifies, the listed checks, a written test record and one 30-minute findings call.

The files must be staged before the window starts. Your team sets up and cleans up the target unless we separately agree to add that work. We don't overwrite production or existing business data. Production recovery and corruption repair aren't included.

Standard target. Test start target: within five business days of target readiness and an agreed booked window. Written report target: two business days after the test.

Priority target. Test start target: within two business days of target readiness. Written report target: two business days after the test. Larger scopes receive explicit dates in the quote.

See when work starts

Discuss a Managed restore test

The test covers one source database, one native backup chain and one isolated target. It includes the defined restore attempt and listed checks. Production recovery, corruption repair and a full DR rehearsal are excluded. We must check the backup chain and isolated target before confirming an exact fixed quote or taking payment.

Priority requires confirmed availability before payment. We’ll confirm whether the service fits, the scope and the dates before payment. This website does not accept payments or diagnostic uploads. See when work starts and what can pause it.

Scope your recovery test Check priority restore-test availability

Tell us what’s happening. Please leave diagnostic evidence out of your enquiry.

What you get

A report your team can work from.

The restore result

See the chain, target, recovery point, timings and success or failure of the agreed attempt.

What the checks show

See the restore result, physical-only integrity checks and customer checks separately. The test does not certify wider application recovery.

What needs attention

Use the result and its dependencies to decide what needs correcting or whether to arrange another test under a separate scope.

Check the scope and deliverables below to see whether this service fits your question.

What is included

What this service covers.

Production recovery during an outage, corruption repair and a full application or disaster recovery rehearsal need a different scope. This test does not guarantee recoverability.

Help me choose another service

One source database on one instance, with allocated source data files totalling no more than 100 GB. Your team supplies one isolated target SQL Server instance, already installed and reachable.
One native SQL backup chain: one full backup, up to one differential and up to 24 log backups, supporting one agreed recovery point. Other backup platforms need a separate quote.
One restore attempt and, if it qualifies, one correction and retry for a procedural error within the same booked time limits. We use a new target database name and don't overwrite production. The test includes one DBCC CHECKDB WITH PHYSICAL_ONLY attempt after a successful restore and up to three approved read-only checks.
Up to three hours of active technical work and four hours elapsed in the agreed test window, plus the standard 30-minute findings call. The files must be staged before the window starts. A larger or slower test needs a quote.

The original backup files remain intact. Use a new target database name, with no overwrite of production or existing business data.

Your test record

The result, timings and validation limits.

The restore result and what the checks prove

You receive a test record with backup identifiers and the chain, target and version information, recovery point, commands or steps, timings, success or failure, validation limits and corrective next steps. It records SQL restore completion, physical-only integrity checks and customer checks separately. Physical-only validation is narrower than full logical integrity checks or business and application validation.

Managed includes the written test record and one 30-minute findings call. Send up to five report questions together within ten business days of delivery for one included round of clarification. This does not include a new test, new recovery point or new investigation.

When is the assessment complete?

We complete the service when we've carried out the agreed restore or test attempt and delivered the evidence and result. A documented failure or timeout can complete the work. If the prerequisites weren't met, that doesn't count as an executed Recovery Proof. We don't promise a successful restore or certify all backups.

A time limit alone does not complete the service. We must document the investigation or executed test, its result and its limits. If the prerequisites aren’t met and the test cannot run, that does not count as an executed test.

What we’ll need

Your environment, the question and access.

Your team supplies the selected chain, target capacity, a compatible SQL version and edition, enough storage, restore permissions, required decryption certificates or keys through an approved channel, and network access. Keys stay in your controlled environment. You must have rights to the data and confirm the target's handling arrangements.
Before booking, we check the backup inventory and header compatibility, agreed recovery point, destination paths, available capacity and realistic duration. The target version must support the source backup; don't assume a backup from a newer version will restore to an older engine. Check features and encryption requirements as well as version numbers. The original backup files must stay intact.
The target must be separate from production and outbound business integrations. Your team sets it up and agrees who will remove the restored data and when. Building a server, buying licences, configuring a backup appliance or exporting the database to BriteThought isn't part of the base package.

Preparing and running the test

Check the chain, target and test window before booking.

We perform the restore and listed checks in the booked window without overwriting production or existing business data. We include one correction and retry only for a clear procedural error, such as an incorrect target path or restore sequence, using the same supplied chain and recovery point. Both attempts are recorded and must fit the three-hour active and four-hour elapsed limits. Corruption repair, rebuilding a chain, another recovery point or a later test needs a new scope.

If an accepted chain fails or the window ends, we preserve the evidence, leave the target in an agreed safe state and report the result. The active-work and elapsed limits don't remove our obligation to report. We estimate feasibility before booking. If our estimate is materially wrong for reasons attributable to BriteThought, we must agree a corrective plan; you aren't forced to buy an add-on.

Completion requires a restore attempt, a recorded result and clear limits. If the restore succeeds but physical-only validation fails, that is a failed validation result, not recovery proven. If validation can't run, the result is restored, validation incomplete. The restore and check timings apply to this test; they don't establish an end-to-end business RTO.

Full CHECKDB, application startup, login and job reconstruction, failover, integration testing, cross-region recovery and forensic analysis aren't included. If needed, they require a separate quote.

Managed only. We check the test requirements first.

Recovery Proof has no Self-Service version or Self-Service upgrade credit. Before payment, we must check the selected backup chain, isolated target and recovery point, confirm whether the test can proceed and give you an exact fixed quote.

Sharing technical evidence

Know what you’re sharing.

What the evidence may contain

Backup chain identifiers and header metadata
Source and target names, versions and paths
Restore commands or steps and test timings
Physical-only check output and approved read-only check results
Business data present on the customer-controlled restore target

Backup files stay in your controlled environment, and the isolated restore target contains business data. Don't upload a full backup through this website. Before the test, agree access, how decryption keys will be handled, target separation, the final target state and who removes the restored data and when.

With Managed, we use named, temporary access through your approved channel, with only the permissions needed. That access is revoked when the work is complete. Permission to read isn't permission to make changes. Don't send credentials, query output, diagnostic packages, backups or sensitive records through general enquiries or ordinary email.

We use your diagnostic evidence to deliver the agreed service. Selling it, using it for public examples or model training, or sending it to external AI services needs separate informed authorisation and suitable agreements.

Use the enquiry form for a brief description of the problem. Keep credentials, query output, diagnostic packages, backups and sensitive records out of the form and ordinary email. Before you share evidence, we'll agree the scope, collection method and privacy, access and retention arrangements.

How long we keep the evidence

Raw packages and temporary working files have a 30-calendar-day retention window after report delivery or cancellation. Final reports are available for 90 days. We keep order and accounting records separately. Before collection, we'll confirm the arrangements and explain any delays deleting backup copies, legal holds or longer retention you've agreed to. This website has no diagnostic upload portal.

When work starts

Turnaround & Priority scheduling

These service targets apply after readiness and availability are confirmed. Enquire for a date; no Priority payment is accepted without a confirmed slot.

Standard test

Start target: within five business days of target readiness and a booked test window. Report target: two business days after the test.

Priority test

Start target: within two business days of target readiness, with confirmed availability. Report target: two business days after the test.

Larger or slower tests

Before booking, we must check the test requirements with you and confirm a fixed quote and specific dates. The base test allows up to three hours of active work within four hours elapsed. All files must be staged before the test window.

When timing starts and pauses

The clock starts only after scope acceptance, payment and required evidence or working access. BriteThought sends a readiness confirmation and dates. An upload receipt is not evidence acceptance.

The clock pauses for missing data, unavailable access, customer approvals, reproduction windows or an agreed third-party dependency. BriteThought notifies you of the reason, required action and revised dates. Pauses are not hidden and do not cover ordinary BriteThought workload.

Intake acknowledgement target: one business day. This acknowledges receipt; it is not completed technical review or an emergency response guarantee.

Hours and availability

Monday through Friday, 9 am to 5 pm America/New_York, excluding US federal holidays observed by the service.

Requests outside service hours enter the next business-day queue unless a separate appointment is confirmed. A symptoms window outside service hours needs a separately agreed appointment.

Priority requires confirmed availability before payment. Payment alone does not create an immediate response commitment. The booking confirmation states the actual reserved start and target delivery date. Priority changes scheduling, not scope, and does not provide remediation, after-hours coverage, continuous emergency support or a resolution deadline. An active outage, inaccessible server or suspected data corruption needs separate incident scoping.

If a confirmed Priority date is missed

If BriteThought cannot honour a confirmed Priority start or report date for reasons within its control, the Priority premium is refunded and the assessment retains its agreed scope. Customer delays and documented external dependencies do not trigger that remedy. Full cancellation and refund are offered when no agreed work has started and no replacement date is accepted. An unconfirmed slot will not be substituted after payment.

Check priority restore-test availability

Outside this service

Production recovery needs a separate scope.

This is a one-time service with a fixed scope. You receive the agreed investigation or test and its written result. Production changes and remediation need a separate agreement, except for the operations expressly included in the isolated Recovery Proof test. We don't guarantee improved performance, uptime, identification of a root cause or recoverability. The service does not certify security or compliance.

This service does not include ongoing monitoring, unlimited revisions, an enterprise audit, compliance certification, licensing advice, an application rewrite, migration, hardware procurement, a production restore, corruption repair, representation in a vendor dispute or routine DBA work. It does not certify security or recoverability.

Why BriteThought

Careful analysis. A report you can use.

Evidence before assumptions

Reports distinguish what was observed, what may be contributing and what remains unknown. Each recommendation explains its evidence, its limits and the next useful test.

A specialist reviews the evidence

A qualified specialist reviews the preparation, restore procedure and recorded results, and is responsible for the engagement. The method sets out the checks, their limits and when further technical review is needed. Participation by a particular founder or engineer isn't guaranteed.

A report your team can use

Use the report with your team, vendor or chosen implementer. It sets out priorities, responsibilities and validation steps so you can decide what happens next. Production changes remain a separate decision.

Common questions

Before we start.

Does a passed test guarantee the next backup will work?

No. It shows what happened with the selected chain on the tested target at that time, within the recorded checks.

Will this recover us during an outage?

This is a scheduled isolated test. Production incident recovery needs a separate assessment.

What remains after the test?

You keep the report and the agreed target state. Your team handles target cleanup unless we explicitly add it to the scope. Agree the cleanup deadline before testing because the restored data is sensitive.

Does Managed include making the recommended changes?

Recovery Proof includes the agreed restore attempt on an isolated target and the listed checks. Production restores, corruption repair and ongoing support are separate.

Does Priority include support around the clock?

Can we share the report and ask follow-up questions?

You can share the test record with your organisation, advisers and chosen implementer. Send up to five questions about the report together within ten business days of delivery; one round of clarification is included. This does not include a new test, recovery point or investigation. Managed includes one 30-minute findings call.

After the report

Decide what happens next.

If a test fails beyond the included procedural retry, correction and a new test need a separate fixed quote. A successful test can inform wider recovery planning; it does not prove full application recovery or that future backups will work.

You can take the report to your team, vendor or another provider. Any further work needs its own clear question and scope. You don’t need to buy another service to have the agreed report explained.

We agree a separate quote for optional follow-up work before it starts. It stays within that scope and does not add ongoing support or automatic charges. Your own team or vendor can act on the report without buying more services.

Useful reading

Tell us what you need to find out.

Tell us what’s happening, how it affects your business and what you want to find out. We’ll check whether the service fits, explain what collection or testing involves and agree when work can start. Please leave credentials and diagnostic evidence out of your enquiry.