<< Back to article

Operation

Operations (Ops) refers to the discipline of deploying, maintaining, monitoring, and ensuring the reliability of computer systems and services. It bridges the gap between software development and production environments, focusing on the practical aspects of running systems at scale.

System Boundaries

Every system operates within boundaries that define its scope, responsibility, and interaction with other systems. Understanding boundaries is essential for designing resilient and maintainable architectures.

See Article - System boundaries for more on defining system boundaries.

Lifecycle of Systems

Systems are born, live, and die. The lifecycle encompasses:

  • Provisioning: Creating and configuring new systems (bare metal, VMs, containers).
  • Deployment: Releasing new versions of software.
  • Operation: Monitoring, maintaining, and patching running systems.
  • Decommissioning: Safely retiring systems and data.

See Article - Lifecycle of Systems for more on system lifecycle management.

System Monitoring

Monitoring is the practice of observing system behaviour to detect issues, understand performance, and ensure service availability. It includes:

  • Metrics: Quantitative measurements like CPU usage, memory consumption, request latency.
  • Logs: Structured or unstructured records of events and errors.
  • Traces: Distributed tracing for understanding request flows across services.
  • Alerting: Automated notifications when metrics exceed thresholds.

See Article - System monitoring for more on monitoring tools and strategies.

Backup, DRP, and DRS

Data and service availability depend on robust backup and disaster recovery planning:

  • Backup: Regular copies of data stored securely, with tested restoration procedures.
  • DRP (Disaster Recovery Plan): Documented procedures for recovering from catastrophic failures.
  • DRS (Disaster Recovery Site): Alternative infrastructure for failover during outages.

See Article - Backup, DRP, DRS for more on backup strategies, disaster recovery, and high availability.