openQA worker should fail with explicit error message if multi-machine test is triggered but requirements are not fulfilled
See #121789. We often have problems to properly setup multi-machine worker configuration and can often only realize by failing openQA tests which is rather expensive. We already have some documentation but we should also have more automatic checks within openQA workers. So for example if an openQA job is triggered requesting the "tap" worker class then the openQA worker can conduct additional checks, e.g. if any parallel worker can actually be reached over network, GRE tunnel, etc., before the actual test code starts and would fail anyway.
- AC1: openQA workers misconfigured as in the example of #121789 do not fail openQA tests (but fail/abort earlier)
- AC2: Non-multi-machine tests are unaffected by multi-machine misconfiguration
- Read what happened in #121789, in particular https://progress.opensuse.org/issues/121789#note-6 and reproduce such setup and change openQA workers to detect such situation itself
#1 Updated by okurz 3 months ago
- Related to action #121789: MultiMachine tests lose ability to communicate added