About this article: This article is reviewed against the cited sources. This is source research, not lab validation. Check the stated versions and your own requirements before making changes.
Direct answer: A usable copy is data—and the metadata to interpret it—that your organisation can restore, query or feed into a replacement environment without the old provider’s proprietary glue. Landing a .zip in customer storage proves possession, not usability. Exit planning and ongoing resilience both improve when you define usable copy as passing a documented restore-to-empty-environment test, not as “the backup job shows green.”
Why it matters
Buyers changing hosters discover too late that they hold opaque disk images without application licences, or CSV dumps without relational context, or exports missing document binaries referenced by database rows. Regulatory enquiries can pose the same question: can you produce intelligible records from what you control?
This article is the portability outcome buyers should demand from exports, backups and documentation together. It deliberately does not re-walk procedures to verify that a backup file arrived via S3 or SFTP—see Verify a backup delivered to S3 or SFTP. The ISV-facing exit angle in Exit and data return when a hosted ISV customer leaves differs from this buyer↔any-hoster view.
NIST SP 800-34 Rev. 1 discusses contingency planning and recovery concepts at programme level; the test design below is practitioner guidance, not text from that guide.
How it works
Layers of “copy” (suggested model, untested):
| Layer | What you have | Usability test |
|---|---|---|
| Possession | Files in customer bucket or tape | Checksum matches manifest |
| Interpretation | Schema, code page, encryption keys | Open in agreed tool; row counts match |
| Reconstitution | DB + binaries + config + identity map | Install clean app; import; login |
| Continuity | Procedures, licences, integrations documented | Run one business task end-to-end |
Define the test once, refresh it annually or after major upgrades (suggested steps—adapt in lab or masked data):
- Scope — which applications, which data classes (production vs archive).
- Empty environment — VM or subscription not sharing production secrets.
- Restore path — use the same artefacts you would rely on if the provider relationship ended.
- Acceptance — named business task on synthetic or masked data where production privacy rules apply (e.g. post a test invoice, run a payroll extract format check, retrieve a sample letter). Block or redirect outbound invoice, payroll, payment and health-related messages so a sandbox acceptance run cannot reach live partners or patients.
- Record — date, owners, gaps found, remediation ticket.
Application semantics dominate usability. A SQL Server backup is restored with vendor-supported RESTORE steps—not merely attached as database files on a mismatched instance. Separately:
- Restore mechanics (Microsoft-supported): a backup cannot be restored to an earlier SQL Server version than the one that created it; restoring an older database onto a newer engine can upgrade it automatically. Encrypted backups need the required encryption keys or certificates accessible on the restore target. See RESTORE (Transact-SQL), Restore a SQL Server Database backup and BACKUP / RESTORE overview (untested here).
- After a successful restore (practitioner guidance): differences in edition features or collation can affect application behaviour even when RESTORE succeeded—validate the business task, not only that the database came online.
Document engine version, key custody, custom modules and job schedules alongside bits on disk.
Customer-held copies complement provider-managed recovery: they reduce negotiation risk at exit and may shorten forensic timelines (practitioner view). They do not automatically replace a tested provider restore if your contract includes one—both can coexist with different purposes.
Encryption keys: if copies are encrypted with keys only the departing provider holds, possession is nominal. Custody of keys must sit with the customer or an agreed escrow for portability to be real.
What to choose and why
Schedule exports for business-readable subsets; retain backups for whole-system rebuild where licensed. Align retention with Australian tax and privacy obligations relevant to your sector—seek qualified advice for legal minima.
Avoid single-format risk: one proprietary appliance format may be fast to restore only on the same stack. Maintain at least one vendor-documented export or backup type your team has tested.
Contract language: ask for artefact types and test cooperation, not vague “data portability.” A provider can reasonably scope effort; a buyer should know what is excluded (e.g. ISV licence media you must obtain directly).
Read next
Related articles you may want next.
- HostingScheduled exports from hosted apps to storage the customer controlsPlan recurring exports to storage your business controls. Specify formats, permissions, validation and retention so exported data remains useful for reporting and portability.Read article
- HostingExit and data return when a hosted ISV customer leavesPlan tenant-scoped exports, acceptance checks, access revocation and an accurate record of active deletion, retained copies and outstanding obligations.Read article
- Backup & RecoveryVerify a Backup Delivered to S3 or SFTP Before RecoveryCheck delivered backup identity, completeness and integrity, then rehearse extraction and recovery using independently available credentials.Read article
Risks and limits
No restore test was executed in lab for this article (untested). LOB and ISV licence terms may forbid running production data in a lab—use masked copies where required.
Does not cover internal backup platform architecture or step-by-step backup-delivery verification.
Sources
- NIST SP 800-34 Rev. 1 — Contingency Planning Guide — contingency and recovery programme concepts (landing summary).
- Microsoft Learn — RESTORE (Transact-SQL) — version restore rules; encryption-key requirements for encrypted backups.
- Microsoft Learn — Restore a SQL Server database backup — RESTORE workflow (not attach-as-files shortcut).
- Microsoft Learn — Back up and restore of SQL Server databases — backup/restore model overview.
- Editorial — usable-copy test layers; post-restore application validation for edition features and collation (practitioner opinion).
Learning notes
- Usable copy means successful business task in a clean environment—not file arrival alone.
- Metadata, versions and keys are part of the copy.
- Possession and reconstitution are different tests.
- Annual refresh catches application upgrades that break restores.
- Portability planning belongs in service onboarding, not the week before exit.
This article was originally published on .