RestorSignal v1.0 / Ops — backup restorability verificationrestorsignal.com
Restorability proven, not assumed

A successful backup
is not yet a successful restore.

Your backup software can tell you a job finished. RestorSignal actually restores the artifact you submit to it, checks the result, measures the operation and keeps an independent proof.

Independent of your backup software — no access to your backup console, reproducible tests, historized proofs.
01 / The gap

A green status does not answer the question that matters.

Backup tools do their job. A job that finished without error is real information — it is not the same thing as a demonstrated restore.

What the backup console tells you

  • The backup job finished.
  • A file was produced.
  • A snapshot is available.
  • A deposit was made on the destination storage.

What it does not demonstrate yet

  • That the artifact is complete.
  • That it is readable and usable.
  • That an actual restore is possible.
  • That a database can actually be reloaded.
  • That the expected data is present.
  • That proof of the test will still exist months from now.
Principle: RestorSignal does not replace your backup tool and does not call it into question. It independently verifies what that green status cannot demonstrate on its own.
02 / Comparison

Two chains. One handoff point.

The backup chain stops at an available artifact. The RestorSignal chain starts exactly where the other one stops.

Classic backup

  • 1. Backup launched ↓
  • 2. Job OK ↓
  • 3. Artifact available

RestorSignal

  • 1. Artifact available ↓
  • 2. Checked ↓
  • 3. Restored ↓
  • 4. Validated ↓
  • 5. Measured ↓
  • 6. Proven
RestorSignal steps in after the backup. It replaces neither your backup tool nor your storage infrastructure.
03 / The progression

Six steps, never merged into a single result.

Each step answers a different question. A higher level that was not run stays explicitly not evaluated — never absorbed into an overall success.

Backed up

The backup job finished, a file exists, a snapshot is available. RestorSignal does not treat this as proof of restorability — it is the starting point, not the conclusion.

Artifact checked

Presence, size, date, fingerprint, rotation and minimal consistency of the artifact are checked before any restore attempt.

Real restore

The content is actually loaded into an environment set up for the test. The MariaDB runner available in v1.0 performs a real import — not a simulation.

Validation

"Restored" does not mean "usable." Technical validation (database opens, tables present, no blocking error) and business/application validation (criteria defined by the client) are two distinct claims.

Measurement

Test start, restore completion, observed duration: every run is timed and logged. RestorSignal reports an observed RTO, never a guaranteed RTO.

Proof

Date, tested artifact, fingerprint, execution environment, results, metrics and verdict become a dated evidence file — not just a memory of a test.

Current v1: the MariaDB runner actually restores and targets level L3. Levels L4 and L5 are never presented as achieved when they have not been run. See the full level definitions →
04 / Independence

Verifying the backup without asking the backup to verify itself.

Producing the archive, choosing what to test, running the check and qualifying the result are four different functions. RestorSignal only handles one: the check and its qualification — never all four at once. Understand the separation of roles →

What RestorSignal does not ask for

  • Veeam admin account
  • PBS account
  • Synology or QNAP access
  • Access to an MSP console
  • Global access to the backup server

What this scope enables

  • Independence from the backup vendor
  • A clear, least-privilege scope
  • Compatibility with several backup chains
  • Comparing restore quality independently of the vendor
The client explicitly submits the artifact to be checked. RestorSignal evaluates that artifact — never the other way around.
05 / Local & Lab

You do not restore a production backup in the middle of production just to check it works.

Two execution philosophies, depending on the client's context.

Local

The test runs inside an environment provided and controlled by the client. The artifact never leaves that environment. Local mode in detail →

Lab

The artifact is voluntarily transferred to a disposable, isolated environment, provisioned for the duration of the test, then destroyed. Lab mode in detail →

RestorSignal does not promise isolation beyond what is actually implemented — the technical detail of each mode is documented and kept current.
06 / Proof as a product

The test result should not disappear with the test.

A historized proof lets you answer, months later, precise questions — not vaguely remember that "we tested it once."

What the file keeps

  • Date and duration of the run
  • Tested artifact and its fingerprint
  • Environment and runner used
  • Checks run and their results
  • Level actually demonstrated
  • What was not tested

Who it serves

  • Operations and periodic review
  • CIOs and IT managers
  • MSPs and end clients
  • Auditors and insurers
  • Business continuity files
We talk about proof, evidence file, verifiable result and history — never certification. RestorSignal is not a certification body.
07 / Use cases

The same gap, in four different contexts.

MSP

"40 client databases every night."

Periodically verify that a dump can actually be reloaded and used — not just that it was produced.

CIO

"My provider tells me it's fine."

The question becomes: when did we last actually restore the backup, and with what result?

Volume

"1.2 TB, no local capacity available."

Lab mode provisions a disposable environment to test without tying up production infrastructure.

MSP → end client

"Something other than a history of green jobs."

Give the client an independent proof of the tests actually performed — not just a table of statuses.

08 / Depth of the check

Not every check answers the same question.

Artifact present → consistent → restorable → usable → verified more deeply. The RestorSignal reference (L1 to L5) separates a check's result from the depth actually demonstrated.

09 / What RestorSignal is not

A definition, not a feature list.

RestorSignal is not

  • A backup solution
  • Backup storage
  • A production orchestrator
  • A replacement for PBS, Veeam, Synology or equivalent
  • A DR promise
  • A contractual RTO/RPO guarantee

RestorSignal is

An independent layer for restore verification and proof retention.

  • Vendor independence
  • Reproducible protocol
  • Checks at variable depth
  • Technical and application validation
  • Measurement and historized proof
RestorSignal — Independent verification

Keep your backups. Add independent proof of restore.

RestorSignal does not replace your backup tool, does not choose your backups, and does not ask for access to your consoles to decide what to test.