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

Control Module FV-01

Fracture Rapid physical disconnection and incident containment

When an incident requires a physical response, Fracture disconnects the registered network crossing at Layer 1. It acts on the link selected in advance, rather than changing a filtering rule while leaving that physical connection available.

Cut the selected path. Keep the rest of the operation in view.

Incident

Path exposed

Outcome

Control taken

The problem

The problem Fracture addresses.

A live attack can exploit an existing connection between environments. Emergency software-policy changes may not be the right final action when responders need that crossing to be physically unavailable.

What it does

When an incident requires a physical response, Fracture disconnects the registered network crossing at Layer 1. It acts on the link selected in advance, rather than changing a filtering rule while leaving that physical connection available.

Before a connection changes

Every deployment starts with a named physical boundary, a defined normal state, mapped dependencies, an agreed decision-maker, failure behaviour and a tested restoration method. Disconnecting one controlled link does not remove alternative routes outside the design.

The physical-control arrangement

A physical circuit-breaker at the assessed crossing. Agreed before it is needed.

A physical circuit-breaker arrangement at the assessed crossing, alongside existing security controls. An approved incident decision initiates the connection-state change; release is a separate authorised action.

  1. Register the crossing

    Register the crossing and the services that rely on it.

  2. Authorise withdrawal

    Authorise incident withdrawal through the agreed decision-maker.

  3. Command the disconnect

    Command the physical disconnect at the registered crossing.

  4. Compare the evidence

    Compare the command, the device-reported state and any independent link verification.

  5. Restore separately

    Restore after separate checks and approval.

What it does

  • Break a selected network path when an authorised incident response calls for physical separation.
  • Acts only on the registered crossings in the agreed design.
  • Records the command, the reported state and any independent verification separately.

What it does not claim

  • Fracture cannot remove an attacker already inside a zone, stop every alternative network route or guarantee unaffected services continue unless their independence has been proven.
  • Act on routes outside the registered design.

See it work

One selected crossing. Four states, kept apart.

Two zones joined by one registered crossing. Follow the crossing from connected, to commanded, to observed disconnected, to independently verified, and then to a separately authorised release.

Affected zone

Zone to protect

Corporate ITWhere the incident is active
OperationsSystems that must stay protected

Stage

Connected

The crossing is in its registered normal state. The services that rely on it are recorded.

Agreed in advance by security and operations.

Commanded
No change.
Reported by the control point
Connected.
Independently verified
Matches the state confirmed at the last test.

Selected boundary

Fracture module emblem: magenta hexagon with a slate blade split awayFracture · FV-01 Registered crossingCorporate IT to operations · Cyber, IT, OT
Normal state
Connected for approved reporting and engineering traffic.
What changes it
An authorised incident decision from the runbook.
What Fracture does
Fracture physically disconnects this crossing at Layer 1 and holds it until a separate release.

Concept illustration. Commanded, reported and verified states are shown separately because they are not the same thing. Disconnecting a crossing does not remove an attacker who is already inside a zone.

Operational boundaries and tests

Agreed in advance. Tested before it counts.

Every deployment starts with a named physical boundary, a defined normal state, mapped dependencies, an agreed decision-maker, failure behaviour and a tested restoration method. Disconnecting one controlled link does not remove alternative routes outside the design.

Before you rely on it

What has to be true first.

  • The design names each boundary, its normal state and who can change it.
  • Dependencies are mapped, including what must keep running.
  • In OT and CNI, the process reaches a safe state before a route is withdrawn.
  • Failure behaviour and the restoration method are agreed and tested.
  • Wireless, cellular and other routes outside the design need their own controls.

Acceptance tests

Prove the operating conditions, not just the command.

  • Command, device-reported state and independent verification are recorded as separate observations.
  • Safe disconnection and restoration are tested in a planned window.
  • The disconnect and the separate release are each tested at the registered crossing.
  • Alternative physical, wireless and cellular paths remain in scope for architecture review.

No guaranteed timing or universal fail-closed state is implied. Observed timings belong to the tested installation.

Related Control Plans

Adopt Fracture through a Control Plan.

The solutions. CP-01 to CP-17 apply the appropriate Modules to defined security and operational problems.

CP-01 Control Plan

CP-01 Ransomware Containment Control Plan iconRansomware Containment

Limit how far ransomware can travel by physically withdrawing the connections it would use, as part of a prepared response.

Open CP-01

CP-02 Control Plan

CP-02 Live Incident Response Control Plan iconLive Incident Response

Withdraw the selected connections an active intrusion is using, with service dependencies assessed before anything changes.

Open CP-02

Industry Blueprints adapt the Plans to each sector's systems, suppliers, continuity requirements and safety constraints. Explore the Blueprints.

Questions

Straight answers.

For security, network and operations owners.

What does Fracture do?

When an incident requires a physical response, Fracture disconnects the registered network crossing at Layer 1. It acts on the link selected in advance, rather than changing a filtering rule while leaving that physical connection available.

What problem does it address?

A live attack can exploit an existing connection between environments. Emergency software-policy changes may not be the right final action when responders need that crossing to be physically unavailable.

How is the physical control arranged?

A physical circuit-breaker arrangement at the assessed crossing, alongside existing security controls. An approved incident decision initiates the connection-state change; release is a separate authorised action.

What does it not claim?

Fracture cannot remove an attacker already inside a zone, stop every alternative network route or guarantee unaffected services continue unless their independence has been proven.

How does Fracture relate to Control Plans?

Modules are the physical control capabilities. Control Plans apply the appropriate Modules to a defined security or operational problem; the design for your own connections is produced with Firevault in Control Studio. Fracture is used in CP-01 Ransomware Containment and CP-02 Live Incident Response.

Is this a validated installation?

No. It is a reference arrangement. Deployment drawings, interlock behaviour, approvals, link-state verification and failure modes are confirmed for your estate in discovery and acceptance testing.

Your next step

Start with one critical connection.

We will help you identify the boundary, assess its dependencies and select the right Modules and Control Plans.

Discuss a critical connectionOr see all nine Modules
    Get started