Firevault, offline secure storage, physically disconnected from the internet
Pricing

BP-10 · Industry Blueprint · Data Centres

Data centres depend on connections. Not all need to stay open.

BP-10 Data Centres Control Blueprint icon

Colocation facilities share power, cooling, fibre and management infrastructure across tenants. Control may govern selected network paths and cross-connect processes, with facility and tenant obligations reviewed together.

  • A support supplier needs a facility-management session without gaining access to unrelated tenant environments. Review the named support and cross-zone boundaries; operationally required links are not assumed disconnectable.
  • Reference design for adaptation with the service owner, not a validated deployment. Confirm complete boundary coverage, operational dependencies and the consequences of opening or closing each selected path.
  • 3 matching Control Plans. Offline Secure Storage® is separate.
  • Control does not replace operational safety systems, network security controls, incident response, supplier governance or service continuity planning. Changing connectivity does not establish a safe process state, stop local execution or guarantee service restoration.

The industry reality

Data Centres: the connections your operation relies on.

Data centres and colocation operations depend on connected systems, specialist suppliers and site infrastructure. A path that is convenient during planned work may also remain available outside that purpose.

Operational interruptionA disrupted data centres and colocation connection may affect services beyond the intended maintenance or response scope.
Safety dependenciesConnections supporting safe operation of data centres and colocation assets require careful dependency review before any intervention.
Persistent third-party pathsAn always-available supplier route can extend access beyond the approved task or maintenance window.
Unclear connection stateLogical policy and tickets alone may not show which selected paths were available at a particular point.

What needs protecting

The service, systems and people behind the connections.

A support supplier needs a facility-management session without gaining access to unrelated tenant environments. Review the named support and cross-zone boundaries; operationally required links are not assumed disconnectable.

Tenant boundariesShared facilities need clear demarcation between tenant networks and facility management systems.
Cross-connect lifecycleAn approved cross-connect may outlive its original business need unless removal and review are explicit.
Facility continuityPower, cooling and environmental systems support site availability and are not assumed disconnectable.

Threats and failure scenarios

Situations to plan for, before the next incident.

A support supplier needs a facility-management session without gaining access to unrelated tenant environments. Control acts on selected physical connections, not every connection in the environment.

Scoped support for Data CentresA support supplier needs a facility-management session without gaining access to unrelated tenant environments. The owner authorises the selected route and its closing conditions.
Review an affected boundaryReview a tenant or shared-management boundary against contracted services, redundant cross-connects and the agreed interruption authority.
CP-12: agree the connection and authorityReview a tenant or shared-management boundary against contracted services, redundant cross-connects and the agreed interruption authority. Use CP-12 only where its scope and approval requirements fit this connection.

Industry architecture

Illustrative sector connections.

A support supplier needs a facility-management session without gaining access to unrelated tenant environments. Review the named support and cross-zone boundaries; operationally required links are not assumed disconnectable.

Illustrative sector connection

Stage

Without Control: the selected routes remain open.

Modules in this Blueprint

Shown when the agreed Control response is selected.

An illustrative ransomware event starts in a connected system and follows the open routes shown here. This is a connection model, not a prediction of a real attack.The open connections allow movement in this simplified model. Actual exposure depends on your environment and existing defences.

Selected connection states

  • Supplier support to facility administrationAlways connected
  • Tenant management to shared-services boundaryAlways connected
  • Tenant production to required cross-connectAlways connected
  • Remote administration to management networkAlways connected
  • Recovery access (OSS)Disconnected
  1. !Illustrative threat starts at · Corporate IT
  2. +Modelled movement to · Supplier access, Identity
  3. +Modelled movement to · Approved access, Tenant networks
  4. +Modelled movement to · Supporting service, Cage and service links, Facility OT
  5. +Modelled movement to · Protected service, Cross-connects

This model only shows the selected routes. An interrupted route does not remove an existing compromise or protect unmodelled paths. Your security tools detect the event; the agreed Control design governs the response. Offline Secure Storage® holds copies separately; it is not a Control module.

Recommended industry Blueprint

Our recommended Blueprint for Data Centres.

A support supplier needs a facility-management session without gaining access to unrelated tenant environments. Control acts on selected physical connections, not every connection in the environment.

Start with these connections
Supplier support to facility administration; Tenant management to shared-services boundary
Agree how access works
Open only for the authorised support task; close under agreed completion conditions
Plan the response
Change only the selected boundary after the service owner approves the impact

Refine this recommendation with tenant service owners, data-centre operations and security leads. Offline Secure Storage® can hold recovery copies separately; it is not a Control module.

As featured in

  • TechRadar Pro logo
  • Yahoo Finance logo
  • Channel Insider logo
  • Security Buyer logo
  • SecurityBrief logo

Operational scenarios

Who decides, what changes, and what is checked.

Agree responsibility and service dependencies before relying on a controlled connection.

  1. Normal operation

    Colocation facilities share power, cooling, fibre and management infrastructure across tenants. Control may govern selected network paths and cross-connect processes, with facility and tenant obligations reviewed together.

  2. Approved work

    A support supplier needs a facility-management session without gaining access to unrelated tenant environments. The owner authorises the selected route and its closing conditions.

  3. Incident response

    A support supplier needs a facility-management session without gaining access to unrelated tenant environments. Control acts on selected physical connections, not every connection in the environment.

  4. Reconnection

    The owner reviews service dependencies and security checks, authorises reconnection and verifies the observed result. Evidence is retained only through the specified supporting systems.

Control does not replace operational safety systems, network security controls, incident response, supplier governance or service continuity planning. Changing connectivity does not establish a safe process state, stop local execution or guarantee service restoration.

Resilience and recovery

Offline Secure Storage®

OSS protects stored business data, recovery copies and retained evidence through physical disconnection. It is a separate storage service, not a Control module or a network containment function.

Control by Firevault governs selected physical connections. Existing detection tools and authorised operators decide when to act. OSS does not detect incidents, command Control or replace operational safety procedures.

Offline Secure Storage (OSS) is separate from Control. If selected, it can hold retained copies of configurations, records or recovery material. Collection, completeness and recovery readiness depend on the supporting tools, retention design and tested recovery procedures.

Explore Offline Secure Storage

Standards and governance

Support your obligations. Keep the boundaries clear.

Alignment may be assessed against ISO/IEC 27001 and applicable data-centre contractual and regulatory requirements; this Blueprint does not imply certification or a compliance outcome. Any assessment needs deployment-specific scope and evidence; this Blueprint provides no certification, complete audit trail or guaranteed response time.

What this addressesA support supplier needs a facility-management session without gaining access to unrelated tenant environments. Control acts on selected physical connections, not every connection in the environment.
What it does not replaceControl does not replace operational safety systems, network security controls, incident response, supplier governance or service continuity planning. Changing connectivity does not establish a safe process state, stop local execution or guarantee service restoration.
Evidence agreed for the deploymentSpecify the requests, approvals, commands, observed states and test results to retain. Logging capability and retention depend on the selected design.

Detection and the decision to act belong to the operational owner and supporting security systems. Identity, approval workflow, session expiry and evidence collection require specified integrations. Execute initiates an approved action; Fracture physically disconnects the selected path through Firebreak hardware. Confirm the observed connection state separately. Anchor keeps critical assets within defined physical connection boundaries; it does not collect or retain records. Transfer governs a designated exchange path, not content inspection or de-identification.

Deployment and proof of concept

Start with one boundary. Test it. Then expand.

Review the first selected connection with tenant service owners, data-centre operations and security leads, agree who can authorise a change and test the service impact.

  1. Map the complete boundary

    Confirm cabling, redundant links and alternate routes. A selected physical cut does not remove paths outside its scope.

  2. Agree an independent command route

    Confirm the available command methods for the selected hardware. Command delivery can be affected by coverage, roaming and handshakes.

  3. Test the effect, not just the command

    Verify the observed link state, service dependencies and failure behaviour in an agreed test window.

Plan your first route

Start with the route you would close first.

Review the first selected connection with tenant service owners, data-centre operations and security leads, agree who can authorise a change and test the service impact.

Outcomes

What changes, and how you will verify it.

Judge the Blueprint on agreed, observed outcomes.

Supplier support to facility administrationOpen only for the authorised support task; close under agreed completion conditions. Agree identity checks, approval, session closure and observed-state checks
Tenant management to shared-services boundaryChange only the selected boundary after the service owner approves the impact. Map alternate routes and preserve services outside the selected scope
Tenant production to required cross-connectNot assumed disconnectable. Safety and continuity owners must assess this dependency separately

Questions worth asking

Before you rely on BP-10.

Practical questions for the people who own these connections.

Colocation facilities share power, cooling, fibre and management infrastructure across tenants. Control may govern selected network paths and cross-connect processes, with facility and tenant obligations reviewed together.

A support supplier needs a facility-management session without gaining access to unrelated tenant environments. Review the named support and cross-zone boundaries; operationally required links are not assumed disconnectable.

Not by assumption. Map dependencies and agree the acceptable service impact with the operational owner before selecting a path.

Control does not replace operational safety systems, network security controls, incident response, supplier governance or service continuity planning. Changing connectivity does not establish a safe process state, stop local execution or guarantee service restoration.

Control by Firevault governs selected physical connections. Existing detection tools and authorised operators decide when to act. OSS does not detect incidents, command Control or replace operational safety procedures.

Review the first selected connection with tenant service owners, data-centre operations and security leads, agree who can authorise a change and test the service impact.

Your next step

Take the next step with data centres.

Review the first selected connection with tenant service owners, data-centre operations and security leads, agree who can authorise a change and test the service impact.

Book a briefing

Prefer to talk first? Speak to the team