Find the revision or container image serving during an investigation
Open the investigation timeline and inspect Serving for runtime observations. Use the service, revision or image identity to select the correct workload.
Read Cloud Run traffic or ECS task evidence#
Timeline follows, with one part for each repository and one for the environment. Serving names each revision that was taking traffic when Proofline read the timeline, its share of traffic and its commit. The numbered entries list deploys, merges, reviews, configuration changes and the alert opening or closing in time order, with the symptom start marked. Each part first shows the five entries nearest the symptom start, plus any other entries cited by the stated likely cause. The link below them opens the rest. A candidate in Where to look and a Likely cause cite these entry numbers.
For an ECS service, Serving names the cluster and service, the container, how many tasks run the image, the image digest and, when build provenance binds it, its commit. A deploy from AWS CodePipeline or Amplify names its CodePipeline execution or Amplify job, and only a succeeded run offers a link to the deploy. When a repository has no successful deploy in the window at or before the symptom start, its part shows its last successful deploy from the 30 days before as Deploy in force at onset. That deploy is not a lead.