Closed andreas-el closed 1 month ago
Should we do an attempt to obtain a FQDN when printing the hostnames?
Should we do an attempt to obtain a FQDN when printing the hostnames?
We could. Currently bjobs will output like this;
(rhel8-bleed-py38-env + bleeding-py38-rhel8) [andrli@st-lintgx0264 komodo]$ bjobs -noheader -o "jobid stat exec_host" 568264
568264 RUN 4*st-rsb17-17-04
Do note that 4*st-rsb17-17-04
is not a valid hostname, but I thought this would be useful to see how many slots were assigned on that specific node.
Do note that
4*st-rsb17-17-04
is not a valid hostname, but I thought this would be useful to see how many slots were assigned on that specific node.
Right. Maybe just better print the string as-is from bjobs to avoid having to parse something we don't really control.
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 91.43%. Comparing base (
8fef8f4
) to head (028a112
).
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
There was an additional comment on the original issue;
While at it, perhaps more realization metadata could be contemplated; RUNPATH, job queue name, LSF log files,
In my opinion the most valuable of these might be the job_queue_name
?
The process of adding this will be identical to what was done for exec_hosts
, even so for the tests, so this should be very low hanging fruit atm.
Issue Resolves https://github.com/equinor/ert/issues/999
git rebase -i main --exec 'pytest tests/ert/unit_tests -n logical -m "not integration_test"'
)When applicable