[CI] investigate unstable scenarios – collect samples of flaky CI runs

[Shastick]

Describe the bug
The CI occasionally fails for no clear reason, and re-running the pipeline again usually shows no issue.

To reproduce
There is no clear way to reproduce this issue (or issues) at the moment

Difference from expected behavior
We would want the CI to consistently fail or succeed.

Possible solution
Currently we re-run flaky CI runs.

Screenshots
If applicable, add screenshots to help explain your problem. Otherwise, remove this section.

System on which behavior was encountered
This happens on the Github CI

Codebase information
Happens on recent main commits, unclear since which commit exactly.

Additional context
As a first step towards finding a fix, we probably want to collect samples of failed CI runs to determine what they share in common, and what scenarios are subject to flakiness.

Observed failures

• on main, fails:
  • FAILURE for "Nominal planning: conflict with higher priority" scenario
  • FAILURE for "ASTM F3411-19 NetRID DSS interoperability" scenario
• on a PR, fails:
  • FAILURE for "Nominal planning: conflict with higher priority" scenario
  • FAILURE for "ASTM F3411-19 NetRID DSS interoperability" scenario
• on a PR
  • FAILURE for "Nominal planning: conflict with higher priority" scenario
  • FAILURE for "ASTM F3411-19 NetRID DSS interoperability" scenario

[Shastick]

Unrelated to the above, I have had a few cases of scenarios failing locally due to clock skew between the mock_uss and the OS, which runs the qualifier. This concerns scenarios that query the interaction log endpoint of the mock_uss.

The details can be found on #750: if anyone runs into intermittent failures caused by missing interaction log entries, this could be an explanation.