feat(eks runway): add runway configuration and CI files for DWS EKS deployment
Summary
This MR adds the necessary Runway configuration files for deploying Duo Workflow Service (DWS) on EKS, following the existing patterns from ai-gateway-eks and duo-workflow-svc configurations.
Changes
-
✅ Created.runway/duo-workflow-svc-eks/default-values.yamlwith base DWS configuration -
✅ Created.runway/duo-workflow-svc-eks/eks-service-staging.yamlfor staging environment -
✅ Created.runway/duo-workflow-svc-eks/eks-service-production.yamlfor production environment -
✅ Updated.gitlab-ci.ymlto include EKS deployment jobs
Configuration Details
The configuration includes:
- Autoscaling: Similar to existing GCP configuration (min: 16, max: 32 for production; min: 1, max: 8 for staging)
- DWS-specific environment variables: All necessary env vars from the Cloud Run configuration
- Resource limits: CPU and memory limits matching the existing DWS deployment
- Health probes: Startup, liveness, and readiness probes configured
- Secrets management: GCP service account credentials and environment variables
References
- Closes #1707
- Runway Kubernetes Getting Started
- AI Gateway EKS deployment: gitlab-com/gl-infra/platform/runway/team#500 (closed)
- AI Gateway EKS epic: gitlab-com/gl-infra/platform/runway&19 (closed)