Feature/job restart policy options #129

[robert-clegg-tessella]

Stories covered: #129

Settings for restarting Spark jobs are configurable in validation_rules.json:

• onFailureRetries
• onFailureRetryInterval
• onSubmissionFailureRetries
• onSubmissionFailureRetryInterval

Also in this PR:

• System tests check for status more frequently to avoid missing completed jobs before they are cleaned up
• python2 script for cleaning up old Docker images

Note also the changes to the wiki guidance on setting TidyFrequency