Instance is not healthy but the deployment shows as succeeded
Issue
In an ECS deployment, the UI shows "successfully deployed" but fails health-check fail 5 seconds later. Health checks are fail but the application still shows as being deployed.The following stage for example is being utilized:
"completeOtherBranchesThenFail": true,
"dependsOn": [],
"executionOptions": {
"successful": true
},
"failPipeline": false,
"name": "Deploy app Stack [dev, ........",
"stageTimeoutMs": 900000,
"type": "deploy"
}
The pipeline completes as succeeded but in the server group logs
there are messages for failures
(service unhealthy).
Cause
Ideally, the deployment should only succeed if the underlying service/task is healthy. This issue is currently being investigated further by Armory Engineering.