journal_check: make sure to select a console to avoid incompletes

[ggardet]

• Related ticket: https://progress.opensuse.org/issues/73297
  Failed test leading to incomplete: https://openqa.opensuse.org/tests/1475717

• VR: https://openqa.opensuse.org/t1479797

[okurz]

sorry, can't say if this is good or not as I don't know in what scope this is normally called. Maybe the "log-console" is used or both sometimes?

@thehejik @jlausuch @mloviska

[ggardet]

Currently, journal_check relies on the console used by previous test, which is the root console.

[jlausuch]

Looks good, but I would appreciate 2 things:

1. Commit message with some more explanation than the title
2. A link to a failed test case and then a VR with CASEDIR which proves the PR works

[ggardet]

1. Commit message with some more explanation than the title

Not sure what additional info you would expect.

2. A link to a failed test case and then a VR with CASEDIR which proves the PR works

Failed test cases are linked in the poo# ticket, but I can add one here. VR are not easy to get as the problem is sporadic.

[Vogtinator]

This is what happens:

1. A test module selects the serial terminal
2. A snapshot is made
3. A module selects a tty console
4. A failure causes a rollback to 2. Now the serial terminal is selected again
5. journal_check calls script_output, which for some reason thinks the tty console is selected (testapi::is_serial_terminal is true)

[2020-11-16T17:50:05.719 CET] [debug] ||| starting journal_check tests/microos/journal_check.pm
[2020-11-16T17:50:05.736 CET] [debug] tests/microos/journal_check.pm:47 called testapi::script_output
[2020-11-16T17:50:05.736 CET] [debug] <<< testapi::type_string(string="cat - > /tmp/scriptWgMmR.sh;\n", max_interval=250, wait_screen_changes=0, wait_still_screen=0, timeout=30, similarity_level=47)
[2020-11-16T17:50:05.737 CET] [debug] <<< consoles::serial_screen::type_string(text="cat - > /tmp/scriptWgMmR.sh;\n", cmd="backend_type_string", max_interval=250, json_cmd_token="AlpatTIZ")
[2020-11-16T17:50:05.738 CET] [debug] tests/microos/journal_check.pm:47 called testapi::script_output
[2020-11-16T17:50:05.738 CET] [debug] <<< testapi::type_string(string="journalctl --no-pager -p err -o short-precise | tail -n +2\n", max_interval=250, wait_screen_changes=0, wait_still_screen=0, timeout=30, similarity_level=47)
[2020-11-16T17:50:05.739 CET] [debug] <<< consoles::serial_screen::type_string(json_cmd_token="ADrMQnmm", max_interval=250, text="journalctl --no-pager -p err -o short-precise | tail -n +2\n", cmd="backend_type_string")
[2020-11-16T17:50:05.739 CET] [debug] tests/microos/journal_check.pm:47 called testapi::script_output
[2020-11-16T17:50:05.739 CET] [debug] <<< testapi::send_key(key="ctrl-d", do_wait=0, wait_screen_change=0)

[Vogtinator]

LGTM, but an updated VR would be nice

[ggardet]

VR passed (failed but as expected), but CI did not start yet...