Skip to main content

Production Deployment

Treat Testudo as a control plane component. Focus on permissions, availability, object storage, remote cluster access, and observability.

  • Run operator, server, and web in a dedicated namespace such as disaster-system.
  • Use the testudo-chart Helm Chart.
  • Keep chart package, values, and image versions aligned.
  • Enable authentication and tracing on server.
  • Send operator logs to centralized logging.
  • Use Events and status as the first troubleshooting entry point.
helm upgrade --install testudo ./testudo-chart-1.0.0.tgz \
  -n disaster-system \
  --create-namespace \
  -f values-prod.yaml

The default public Docker Hub images do not require an image pull Secret. If you use a private registry, configure the pull Secret and image repositories in your private values file:

imagePullSecret:
  existingSecret: default-secret

images:
  operator:
    repository: registry.example.com/softcdata/testudo-operator
  server:
    repository: registry.example.com/softcdata/testudo-server
  web:
    repository: registry.example.com/softcdata/testudo-web

Availability​

  • Server can run with multiple replicas.
  • Operator should use leader election.
  • Object storage should be highly available.
  • Remote cluster credentials should be rotated.
  • disaster-server should normally remain ClusterIP.

Upgrade​

  1. Back up values and important CRD instances.
  2. Confirm chart and image versions.
  3. Run helm template and dry-run upgrade.
  4. Avoid upgrading during long-running operations.
  5. Run helm upgrade.
  6. Watch instances, operations, and task events.