[3006.x] Fix test_system flaky setup_teardown

[m-czernek]

What does this PR do?

I noticed that when running the testsuite, we intermittently get the following error:

04:28:52  _________________ ERROR at teardown of test_set_computer_desc __________________
04:28:52  Traceback (most recent call last):
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/_pytest/runner.py", line 341, in from_call
04:28:52  result: Optional[TResult] = func()
04:28:52  ^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/_pytest/runner.py", line 262, in <lambda>
04:28:52  lambda: ihook(item=item, **kwds), when=when, reraise=reraise
04:28:52  ^^^^^^^^^^^^^^^^^^^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/pluggy/_hooks.py", line 265, in __call__
04:28:52  return self._hookexec(self.name, self.get_hookimpls(), kwargs, firstresult)
04:28:52  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/pluggy/_manager.py", line 80, in _hookexec
04:28:52  return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
04:28:52  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/pluggy/_callers.py", line 60, in _multicall
04:28:52  return outcome.get_result()
04:28:52  ^^^^^^^^^^^^^^^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/pluggy/_result.py", line 60, in get_result
04:28:52  raise ex[1].with_traceback(ex[2])
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/pluggy/_callers.py", line 39, in _multicall
04:28:52  res = hook_impl.function(*args)
04:28:52  ^^^^^^^^^^^^^^^^^^^^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/_pytest/runner.py", line 182, in pytest_runtest_teardown
04:28:52  item.session._setupstate.teardown_exact(nextitem)
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/_pytest/runner.py", line 537, in teardown_exact
04:28:52  raise exceptions[0]
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/_pytest/runner.py", line 526, in teardown_exact
04:28:52  fin()
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/_pytest/fixtures.py", line 701, in <lambda>
04:28:52  subrequest.node.addfinalizer(lambda: fixturedef.finish(request=subrequest))
04:28:52  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/_pytest/fixtures.py", line 1031, in finish
04:28:52  raise exc
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/_pytest/fixtures.py", line 1024, in finish
04:28:52  func()
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/_pytest/fixtures.py", line 911, in _teardown_yield_fixture
04:28:52  next(it)
04:28:52  File "/usr/lib/venv-salt-minion/lib/python3.11/site-packages/salt-testsuite/tests/pytests/functional/modules/test_system.py", line 83, in setup_teardown_vars
04:28:52  res = service.start("systemd-timesyncd")
04:28:52  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/salt/loader/lazy.py", line 149, in __call__
04:28:52  return self.loader.run(run_func, *args, **kwargs)
04:28:52  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/salt/loader/lazy.py", line 1234, in run
04:28:52  return self._last_context.run(self._run_as, _func_or_method, *args, **kwargs)
04:28:52  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib64/python3.11/site-packages/salt/loader/lazy.py", line 1249, in _run_as
04:28:52  return _func_or_method(*args, **kwargs)
04:28:52  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
04:28:52  File "/usr/lib/venv-salt-minion/lib/python3.11/site-packages/salt/modules/systemd_service.py", line 872, in start
04:28:52  raise CommandExecutionError(_strip_scope(ret["stderr"]))
04:28:52  salt.exceptions.CommandExecutionError: Job for systemd-timesyncd.service failed.
04:28:52  See "systemctl  status systemd-timesyncd.service" and "journalctl  -xe" for details.
04:28:52  ---------------------------- Captured log teardown -----------------------------
04:28:52  ERROR    tests.pytests.functional.conftest.loaded.int.module.cmdmod:impl.py:295 Command '/usr/bin/systemd-run' failed with return code: 1
04:28:52  ERROR    tests.pytests.functional.conftest.loaded.int.module.cmdmod:impl.py:295 stderr: Running scope as unit: run-r0c01f3d9cc894f94aa5db15d337abb62.scope
04:28:52  Job for systemd-timesyncd.service failed.
04:28:52  See "systemctl  status systemd-timesyncd.service" and "journalctl  -xe" for details.
04:28:52  ERROR    tests.pytests.functional.conftest.loaded.int.module.cmdmod:impl.py:295 retcode: 1

The reason for that is the service start request repeated too quickly, refusing to start error, which seems to happen when restarting a service more than 5 times in a 10s window as per [0].

This diff should fix the test flakiness; when service.start throws an exception, we wait 10s and try again. If the service start still fails, we fail. Otherwise, we continue.

[0] https://serverfault.com/questions/845471/service-start-request-repeated-too-quickly-refusing-to-start-limit

Merge requirements satisfied?

[NOTICE] Bug fixes or features added to Salt require tests.

• Docs
• Changelog - https://docs.saltproject.io/en/master/topics/development/changelog.html
• Tests written/updated

Commits signed with GPG?

Yes

Please review Salt's Contributing Guide for best practices, including the
PR Guidelines.

See GitHub's page on GPG signing for more information about signing commits with GPG.

[welcome]

Congratulations on your first PR being merged! 🎉