Azure-Samples / Azure-Samples/tutor
P0: Deployment blocked on ACA provisioning timeout in azd workflow
- Dominant language
- Python
- Stars
- 17
- Forks
- 8
- PR merge metrics
- No merged PRs in 30d
Description
## Incident
Deployments to main are repeatedly blocked in .github/workflows/azd-deploy.yml during provision -> Provision infrastructure.
## Evidence
- Failed runs: 23086972724, 23080475978, 23078568547, 23077617849, 23076942398, 23073936361, 23073800290, 23073706881, 23073405755, 23072504072, 23071911633, 23071628689, 23071177921
- Latest failure log signal: zurerm_container_app.backend_services["questions"] stayed creating >20m; ssays started; then workflow hit ttempt exceeded 45m timeout and exited 1.
- AzureContainerAppsSpecialist findings: multiple apps in utor-dev reached provisioningState=Failed with ContainerAppOperationError / Operation expired.
## Architecture context
- ADR-003 (ACA microservices) expects independent service deployment behavior.
- ADR-004 (Terraform AVM) expects stable declarative provisioning.
Current timeout chokepoint blocks all downstream deploy jobs.
## Acceptance criteria
- [ ] Confirm root cause at ACA control-plane/app revision level for lms-gateway, questions, ssays.
- [ ] Capture per-app failure timeline and correlation IDs from Activity Log.
- [ ] Define a mitigation that unblocks deploys without bypassing GitHub workflow deployment policy.
- [ ] Link follow-up implementation issues and close this issue only after first green deploy.
Contributor guide
Assessment
This issue has not been assessed yet.