Skip to content

System Limits

Suite Cleanup Timeout

Suite cleanup workflows have a 45-minute timeout. This limit applies everywhere:

  • Auto mode (rocketship run -af)
  • Manual mode (rocketship start server / rocketship stop server)
  • Rocketship Cloud (persistent servers)

If cleanup exceeds 45 minutes, Temporal cancels the workflow. The run outcome (passed/failed) is NOT changed - cleanup failures are logged but don't fail the suite.

Why This Matters

Cleanup can include long-running operations:

  • delay steps: { duration: "10m" }
  • Environment teardown scripts: ./env/down.sh (30+ minutes)
  • Cloud resource deletion: RDS, CloudFormation, Kubernetes clusters

Best Practices

Prioritize critical cleanup first: 

cleanup:
  always:
    - name: "Delete test database"  # Critical - do first
      plugin: supabase
      config: { action: execute_sql, query: "DROP DATABASE test_db" }
    - name: "Delete logs"  # Nice to have - do last
      plugin: http
      config: { method: DELETE, url: "{{ logs_url }}" }

Use idempotent operations: Design cleanup to be safely re-runnable if it times out or fails partway through.

Break up long operations: Instead of one 30-minute script, use multiple smaller steps that can fail independently: 

cleanup:
  always:
    - name: "Stop services"
      plugin: script
      config: { cmd: "./scripts/stop-services.sh" }
    - name: "Delete volumes"
      plugin: script
      config: { cmd: "./scripts/delete-volumes.sh" }
    - name: "Remove network"
      plugin: script
      config: { cmd: "./scripts/remove-network.sh" }

Log progress: Add log steps between long operations so you can see how far cleanup progressed: 

cleanup:
  always:
    - name: "Starting teardown"
      plugin: log
      config: { message: "Tearing down infrastructure..." }
    - name: "Delete stack"
      plugin: script
      config: { cmd: "./env/down.sh" }
    - name: "Teardown complete"
      plugin: log
      config: { message: "Infrastructure torn down" }