Kanban for DevOps and Infrastructure Teams

Kanban for DevOps

DevOps teams handle a mix of planned work (infrastructure improvements, automation) and unplanned work (incidents, urgent requests). Kanban's continuous flow model is ideal for this dual-track workload.

Recommended Column Layout

Backlog → Ready → In Progress → Testing → Staging → Deployed → Monitoring

• Backlog -- Infrastructure improvements, automation tasks, tech debt.
• Ready -- Items with clear requirements and runbooks.
• In Progress -- Active work. WIP limit: team size x 1.
• Testing -- Validation in a non-production environment.
• Staging -- Deployed to staging, awaiting final smoke tests.
• Deployed -- Pushed to production.
• Monitoring -- Observing for 24-48 hours before considering it done.

Incident Swimlane

Add a horizontal swimlane (or separate color) for incidents:

• Red cards -- Active incidents. These bypass WIP limits and take priority.
• When an incident is raised, create a red card and move it directly to In Progress.
• After resolution, move through Testing and Monitoring before closing.

Color Labels for DevOps

• Red -- Incident / outage
• Orange -- Security patch / vulnerability
• Blue -- Infrastructure improvement
• Green -- Automation / CI-CD improvement
• Yellow -- Monitoring / observability
• Purple -- Documentation / runbook

Change Management

For infrastructure changes, include in the card description:

• Change request ID
• Rollback plan
• Blast radius (which services are affected)
• Scheduled maintenance window

Metrics That Matter

• Mean Time to Recovery (MTTR) -- How quickly incidents are resolved. Track as cycle time for red cards.
• Change Failure Rate -- Percentage of deployments causing incidents. Track by counting red cards created within 24 hours of a deployment.
• Deployment Frequency -- Cards reaching Deployed per week.