remove stale daemon pid in status check

[unkcpz]

The problem (#2795) that stale daemon PID not found after reboot when daemon is not stop correctly also occurred in command verdi daemon status and verdi status.

Also calling function delete_stale_pid_file in verdi status check.

Or maybe it is better to print tips on how to solve the problem in get_daemon_status rather than remove files automatically?

[coveralls]

Coverage increased (+0.008%) to 74.354% when pulling 3677b18 on unkcpz:daemon-stale into 406dbb4 on aiidateam:develop.

[sphuber]

Or maybe it is better to print tips on how to solve the problem in get_daemon_status rather than remove files automatically?

The problem with this approach is that it is difficult to determine the source of the problem when a call to daemon times out. Note that the delete_stale_pid_file should only remove the pid file if the corresponding process can no longer be found. This should be fine in principle. Did you find a situation where the pid file was removed unjustly? If not, why do you think we should remove these lines?

[unkcpz]

I didn't find situation where the pid file was removed unjustly.

My PR suggests also adding delete_stale_pid_file in verdi status and verdi daemon status.

[sphuber]

Ah you are right, I glossed fast over the diff and interpreted it exactly the other way around.

[unkcpz]

I think this is not an urgent PR, but it can be confusing for users, especially beginners.

[giovannipizzi]

Thanks a lot @unkcpz!

In the long term, if we realise we are starting to add this in too many places, we should probably find a better way to avoid putting the call in many places. But for now this a useful fix for user and we approve it.