At the moment we only report if a Testground task is scheduled, processing or completed.

• We should include an exitcode for the task run in the task schema and display it on the dashboard
• When we get a failed run (see definition of fail below), testground should fetch extra information such as:
  a) number of outcome: ok events / vs total instances per composition group
  b) testground should inspect runner env events, for example in cluster:k8s - ErrImagePull, or OOM, or instances blocked in Pending state (which would result in global timeout for the run) and attach them to the task in case of failures.

Definition of exitcode:

1. If a task failed to compile, the exitcode should be >0 and task should be marked as failed.
2. If a task compiles and its run completes, and all instances are succeeded and all logs include outcome: ok event, then the run is successful.
3. For now, every other case, means a run failed.

Note that partial successes are not in scope for this iteration. In the future we might want to split the failed runs into more categories, and include partial success (for example a test plan with 1000 instances, which are mostly independent and 999 manage to succeed with outcome: ok event). This is why it is important to also include extra information as part of runs that are not 100% successful.