When a build fails, a notification is sent to an SNS topic and forwarded to the WatchbotNotificationEmail that was provided when the ecs-conex stack was created. A notification will look similar to this:
At Tue, 26 Jul 2016 23:29:50 GMT, processing message a7492004-8ca8-4322-9299-2e82bb649163 failed on ecs-conex-production
Task outcome: delete & notify
Task stopped reason: Essential container in task exited
Message information:
MessageId: a7492004-8ca8-4322-9299-2e82bb649163
Subject: webhook
Message: {"ref":"refs/heads/test-branch","after":"81e48385715d60cae6f6d9ae818d8148590a9902","before":"c2abf76a55709b2f5eb27eeb1c0d33d4408ea963","repository":{"name":"ecs-conex","owner":{"name":"mapbox"}},"pusher":{"name":"rclark"}}
SentTimestamp: 1469575768248
ApproximateFirstReceiveTimestamp: 1469575768250
ApproximateReceiveCount: 1
Runtime resources:
Cluster ARN: arn:aws:ecs:us-east-1:123456789012:cluster/ecs-cluster-production
Instance ARN: arn:aws:ecs:us-east-1:123456789012:container-instance/2e14b317-0909-4ecc-ab88-d94fe64d2167
Task ARN: arn:aws:ecs:us-east-1:123456789012:task/798b49eb-49d7-4abb-a305-82a6e723caf6
First off all, check the Message JSON to help identify the commit that caused a failure, the repository that was being built, and the person who was responsible for the commit.
When used in conjunction with ecs-watchbot, logs from ecs-conex containers will be written to a CloudWatch Log Group. Watchbot creates this resource with the name <stackName>-<stackRegion>-<serviceVersion>. Within your log group, individual message logs will contain the MessageId — a7492004-8ca8-4322-9299-2e82bb649163 in this example — in the log stream.
If there are more questions, the Runtime resources indicate the ECS cluster, the EC2 instance, and the ECS task that attempted the build. You can use these for closer inspection via further ECS API requests.