Batch job submissions that fail with Batch.AWSBatchException are not retried

[asjohnston-asf]

This is a rare case of a transient error from the AWS Batch API not being correctly retried.

Job 713ee4d5-73cd-4cdd-b1a6-e18d2ef8839b received the following error when the step function attempted to submit the INSAR_GAMMA job to AWS Batch:

{
  "resourceType": "batch",
  "resource": "submitJob.sync",
  "error": "Batch.AWSBatchException",
  "cause": "null (Service: AWSBatch; Status Code: 500; Error Code: AuthorizerConfigurationException; Request ID: 9db2e8a7-90d7-43fc-95e2-80842395bb3b; Proxy: null)"
}

The step function is configured to only retry Batch.ServerException:
https://github.com/ASFHyP3/hyp3/blob/develop/apps/step-function.json.j2#L115

In this particular case the step function failed to complete since there was no batch log file to be processed by the JobFailed step. From the user's perspective, the job will appear in a permanent RUNNING status.