Our Jenkins stress test jobs configure test/format to run for a certain length of time. test/format runs mostly conform to the configured runtime, but occasionally the runtime significantly exceeds the configured max. In those cases it would be valuable to track which configuration caused the slow test cycle.

We could do that by:

• Tracking runtime in Jenkins stress testing, and failing after a long iteration (this would probably require switching from single command multi-run test/format jobs, to multiple runs of single iterations).
• Noticing unexpectedly long runtimes within test/format and exiting with a failure
• Noticing unexpectedly long runtimes within test/format and printing the configuration. Not preferred, since we will probably ignore the information.
• Your good idea goes here.