CFTS Documentation

Monitoring and Automation

Monitoring and automation are closely connected within the CFTS operating environment.

Monitoring provides evidence about what is happening. Automation can then help present that information, perform repeatable checks, reduce manual work, or carry out a controlled action where this is appropriate.

CFTS does not treat automation as a substitute for understanding the infrastructure. Automated actions should be based on clear conditions, observable results, and a known recovery path.

Internet and Service Monitoring

CFTS operates multiple internet and service paths. Monitoring is used to provide visibility of availability, latency, service condition, and where appropriate the state of different uplinks.

This information can support:

  • fault diagnosis
  • identification of provider or path problems
  • failover decisions
  • service-status reporting
  • capacity and trend review
  • verification after a change

Selected information may be exposed through client-facing status services such as isp-status.cfts.co/view, while more detailed operational information remains available to CFTS engineers.

Failover and Resilience

A failover mechanism is only useful if it can distinguish between a real service failure and a temporary or local condition.

For this reason, CFTS may combine monitoring, health checks, routing information, and service-specific tests before an automated or assisted change is made.

Depending on the environment, failover may be:

  • automatic
  • policy-based
  • assisted by monitoring
  • manually controlled with automated verification

The correct approach depends on the service and the consequences of an incorrect change.

Diagnostics

Purpose-built diagnostics can reduce the time needed to answer basic operational questions.

Rather than requiring an engineer to connect independently to several systems, diagnostic tools may collect or present information such as:

  • service reachability
  • host state
  • DNS results
  • network path information
  • uplink condition
  • response timing
  • selected platform health

These tools are intended to assist engineering judgement, not replace it.

Controlled Automation

CFTS favours bounded automation over unrestricted control.

Where software performs operational actions, the preferred model is to expose only the specific functions required for the task rather than provide a general-purpose command interface.

This helps make actions easier to audit, test, secure, and support.

Examples may include:

  • restarting an approved service
  • collecting a defined diagnostic set
  • checking a host or service state
  • updating an approved configuration value
  • triggering a documented workflow
  • carrying out a defined VM operation

The exact level of automation depends on the risk of the action and the environment in which it operates.

AI-Assisted Automation

CFTS is also developing locally controlled AI-assisted operational capabilities.

The intention is to use AI as an additional interface to documentation, diagnostics, and approved tools rather than provide unrestricted access to infrastructure.

Where AI is allowed to interact with operational systems, the same principle applies: actions should be narrowly defined, controlled, and appropriate to the task.