DISASTER RECOVERY

When the region is gone, you recover. Not discover that half of prod was never in code.

Facets brings your entire running estate under one governed blueprint in days, not months: 100% of everything you connect us to, with the gap report showing exactly what is left to reach. Recovery you can rehearse, from the estate you already run. No re-platforming. AWS, GCP, and Azure.
Days
Not months to full coverage
100%
Of everything you connect us to
3
Clouds, one blueprint
Zero
Re-platforming
AgNext
AskMyGuru
Aster
Astuto
Capillary
CoinSwitch
CommerceIQ
Fortune 50 Automotive
FourKites
GGX
INTEGRTR
MoveInSync
MPL
Niyo
PubMatic
Purplle
Refold AI
Rupicard
SaaS Labs
SimpliSmart
Snabbit
Spectrum
Tekion
Treebo
Vymo
ZeonAI Labs
Zyla
AgNext
AskMyGuru
Aster
Astuto
Capillary
CoinSwitch
CommerceIQ
Fortune 50 Automotive
FourKites
GGX
INTEGRTR
MoveInSync
MPL
Niyo
PubMatic
Purplle
Refold AI
Rupicard
SaaS Labs
SimpliSmart
Snabbit
Spectrum
Tekion
Treebo
Vymo
ZeonAI Labs
Zyla
THE PROBLEM

Your DR plan is a document you will read for the first time during the outage

And half of production was never in code

Runbooks that rot

The plan was accurate the day it was written and drifts a little every day after.

Un-codified production

The click-ops glue, the secrets, the wiring and the dashboards were never captured anywhere.

Recovery you cannot rehearse

You find out what the plan missed at the worst possible moment, with the clock running.

The common worst case is not lost backups. It is that production was never fully infrastructure as code, so there is nothing complete to recover from.

THE LANDSCAPE

Everyone recovers the machines. Nobody governs the whole product.

Backups restore machines. Recovery tools replay captured state. None of them is a governed model that also runs your production.

What the rest of the market does

Backups (Veeam, Druva): restore VM images and data. Networking, IAM and wiring are not captured in the backup, so you redefine them in the recovery plan.
Replication (Zerto, Azure Site Recovery): mirror running machines into a target region. Compute must already exist, and networking is limited to mapping or recreating VNets and subnets.
Recovery-as-code tools (Firefly, ControlMonkey): snapshot your cloud into Terraform and replay it to restore infrastructure. They recover captured state, not a governed model your teams operate from every day.
Cloud rebuild tools (Commvault Cloud Rewind, Cohesity Cloud Rebuild): orchestrate a rebuild from infrastructure as code at restore time. Cohesity needs your existing Terraform and targets select AWS workloads. Commvault auto-discovers AWS, Azure or GCP.

What Facets does

Discovers the complete running estate, including the click-ops nobody codified, across clouds and external services like MongoDB Atlas.
Pins it into one continuously governed blueprint that does not silently rot between drills.
Is the same model that operates production, so recovery is a property of how you run, not a separate restore job.
Rebuilds the networking, IAM, config, secrets and wiring, and tells you exactly what it could not reach.
# The same blueprint that runs prod is the recovery artifact
network:        { type: vpc }
gke:            { type: kubernetes_cluster }
mysql:          { type: managed_database }
messaging:      { type: pubsub }       # the eventing fabric backups miss
mongodb_atlas:  { type: external }     # discovered off-cloud, pinned too

# One governed blueprint. One human-gated action spins a cold DR region.
HOW IT WORKS

Recovery you can rehearse, from the estate you already run

Honest throughout: recovery time is spin plus restore, and data restores from your existing backups.

1

Discover and import

Read-only discovery of the complete running estate.

  • Named read-only scope: GCP roles/viewer, AWS ReadOnlyAccess, Azure Reader
  • Finds the un-codified click-ops, not just what is already in IaC
  • Imports at zero change, gated on a plan with no replace and no destroy
2

Pin one governed blueprint

The complete estate becomes a single versioned artifact.

  • Networking, IAM, config, secrets references, wiring and observability
  • One blueprint across AWS, GCP, and Azure
  • Exportable as standard Terraform, so you can exit the platform at any time
3

Keep it drift-honest

Continuous re-discovery prevents the blueprint from silently rotting.

  • Dated, approved drift records instead of a document that ages
  • A coverage report that shows exactly what is and is not captured
  • Gaps are known, not silent, with a reason on every line
4

Spin a cold DR region

One human-gated action rebuilds the whole product.

  • Recovery runs inside your own DR project, production untouched
  • Recovery time is honest: spin plus restore
  • Data restores from your existing backups, no rip and replace

One anonymized production estate, one governed blueprint

A growth-stage SaaS communications platform, messy not big: a fraction was in IaC, with default service accounts and no central pipeline. The kind of estate every team actually runs. We imported it into a single Facets blueprint.

1,100+
Production resources under one blueprint
1,000+
Pub/Sub topics and subscriptions, the eventing fabric backups miss
Off-cloud
MongoDB Atlas cluster discovered and pinned alongside the GCP estate

What backups would have missed

Not just the VMs and the database

Beyond compute and data, the blueprint captured the messaging fabric, managed caches, task queues, object storage and the network itself. Backups and replication never see most of it.

Across providers, not just one cloud

An external MongoDB Atlas cluster was discovered and pinned next to the GCP estate. Recovery is not cloud-locked, and there is no re-platforming.

Gaps known, not silent

Where a resource cannot be reached, the report says so and why. Coverage climbs toward 100% as you connect what is left. The honest gap report is the trust signal.

Honest scope

This proves Facets can see and govern the complete estate. We have not published a measured end-to-end recovery time. Recovery is honest spin plus restore, and your first drill from the artifact sets your real number.

TRUST BY DESIGN

The questions that actually decide this

Read-only access, honest secrets handling, and a blueprint you can run without us

Read-only by design

Discovery never writes.

  • Named scopes: roles/viewer, ReadOnlyAccess, Reader
  • Permission set reviewable before you grant it
  • Recovery writes only in your DR project
  • Human-gated, with production untouched

Secrets, handled honestly

Reference-only by default.

  • We record that a secret exists, never its value
  • Material never leaves your boundary
  • KMS and HSM backed references
  • Control document and SOC 2 on request

Export to exit, no lock-in

Your infrastructure stays portable.

  • Export your estate as standard Terraform
  • Take it with you to exit the platform anytime
  • Readable, reviewable infrastructure as code
  • No proprietary format to depend on
SEE IT

Disaster recovery, without re-platforming

Recovery you can rehearse, from the estate you already run. Two minutes, calm and honest throughout.

1

Disaster recovery, without re-platforming

How Facets discovers the estate you already run, pins it as a versioned blueprint, keeps it drift-honest, and spins a cold DR region with one human-gated action.

  • Discover the complete estate, not just infra
  • One governed blueprint as the recovery artifact
REGULATED BUYERS

A DORA, NIS2 and ISO 22301 readiness track, mapped not just claimed

For teams that have to take continuity to a board and an auditor.

DORA control map

Business impact analysis and at-least-annual continuity testing (Article 11), plus backup, restoration and RTO/RPO recovery objectives with redundant ICT capacity where required (Article 12), mapped to the blueprint. DORA applies since 17 January 2025.

Drill records as audit evidence

Tamper-evident, timestamped records of every rehearsed recovery, ready for an auditor.

Secrets handling statement

Reference-only, KMS and HSM backed, with a control document available for procurement review.

EU data residency

Blueprint metadata can be kept in region for data residency requirements.

Exit on your terms

Export your estate as standard Terraform to leave the platform at any time, which answers the third-party-risk and lock-in question directly.

A regulated peer, under NDA

Request a private reference call rather than a logo wall. We do not fabricate proof.

Facets can operate production too. Your call.

The same blueprint can run production

The artifact that recovers you can also operate your estate day to day. That is what keeps it real rather than a document that rots beside production.

Disaster recovery does not require it

Adopting DR is a read-only assessment. Choosing to have Facets operate production is a separate, deliberate, separately-assessed decision.

And you can always exit

Whatever you choose, Facets exports your estate as standard Terraform, so you can leave the platform at any time. No lock-in either way.

Virender Bisht
When we needed air-gapped deployments for our partner bank, Facets built a solution for that as well on short notice. Brilliantly done and nicely executed.
Read Case Study

Virender Bisht

Co-Founder & CTO · Niyo

Niyo logo

Get a read-only DR coverage assessment

A free, honest map of your drift and IAM rot, even if you never buy DR. The report ends with a scoped drill you can run from the artifact.

Frequently asked questions