Support ECS Monitoring containers alongside deploy task.

[iBotPeaches]

If you use ECS Monitoring - a sidebar monitoring container is launched alongside your containers. This means when we check for the exit code we look at the first container. We have many in a non-consistent order.

{
  "containers": [
    {
      "containerArn": "arn:aws:ecs:us-east-1:xxx:container/tg-alpha/xxx/267357fc-b146-44a0-a463-cd0e758299c1",
      "taskArn": "arn:aws:ecs:us-east-1:xxx:task/tg-alpha/xxx",
      "name": "node",
      "image": "xxx.dkr.ecr.us-east-1.amazonaws.com/node-alpha:v0.0.0-6-gcb02-A-TG-1",
      "imageDigest": "sha256:xxx",
      "runtimeId": "xxx-2982235661",
      "lastStatus": "STOPPED",
      "exitCode": 1,
      "networkBindings": [],
      "networkInterfaces": [
        {
          "attachmentId": "067c937a-83f0-47b1-a3d6-8a9849eac063",
          "privateIpv4Address": "10.0.3.145"
        }
      ],
      "healthStatus": "UNKNOWN",
      "cpu": "0"
    },
    {
      "containerArn": "arn:aws:ecs:us-east-1:xxx:container/tg-alpha/xxx/e37ca100-622b-4c2f-b5a3-c3792254a229",
      "taskArn": "arn:aws:ecs:us-east-1:xxx:task/tg-alpha/xxx",
      "name": "aws-guardduty-agent-i2erCm",
      "imageDigest": "sha256:xxx",
      "runtimeId": "xxx-2715164973",
      "lastStatus": "STOPPED",
      "exitCode": 0,
      "networkBindings": [],
      "networkInterfaces": [
        {
          "attachmentId": "067c937a-83f0-47b1-a3d6-8a9849eac063",
          "privateIpv4Address": "10.0.3.145"
        }
      ],
      "healthStatus": "UNKNOWN"
    }
  ]
}

Sometimes we check the wrong container depending on the order we get. In this sample we have the sidecar guard duty and our migration container. One failed (exit code 1), other failed gracefully (stopped - 0).

We have a few options.

1. Introduce a command to ask for the container name to check (so we filter).
2. Extract the container name from the prepare task image changes to reduce manual container name input.
3. Check all containers on prepare job for zero exit.