- 28 Jun, 2017 2 commits
-
-
Tim Wickberg authored
Unhandled return code makes Coverity sad, use the macro for consistency in these cases and to silence the warnings.
-
Tim Wickberg authored
Coverity noticed we don't consistently use it and started flagging these. Switch to the macro for consistency.
-
- 27 Jun, 2017 14 commits
-
-
Morris Jette authored
-
Morris Jette authored
The re-use of error_code results in the function returning an error when it really functions properly. Added new variable to use.
-
Tim Wickberg authored
Use part_is_visible instead, which does not modify the partitions (and thus require the part write lock).
-
Tim Wickberg authored
No longer require part write lock.
-
Tim Wickberg authored
They now only need the part read lock, not the write lock.
-
Tim Wickberg authored
Only needs part read lock, whereas part_filter_set/clear require the part write lock which limits concurrency. Convert one location in pack_all_part() while here.
-
Tim Wickberg authored
Also remove long-hidden test program from parse_time.c.
-
Tim Wickberg authored
-
Tim Wickberg authored
-
Tim Wickberg authored
-
Tim Wickberg authored
Fix one location in sacctmgr that was altering the value to use a heap-allocated value instead.
-
Tim Wickberg authored
Previous implementation is not thread-safe.
-
Tim Wickberg authored
Replace with direct references to the struct.
-
Tim Wickberg authored
-
- 26 Jun, 2017 2 commits
-
-
Morris Jette authored
-
Dominik Bartkiewicz authored
Improve backfill scheduling algorithm with respect to starting jobs as soon as possible while avoiding advanced reservations. bug 3757
-
- 24 Jun, 2017 4 commits
-
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
- 23 Jun, 2017 7 commits
-
-
Morris Jette authored
bug 3886
-
Morris Jette authored
-
Tim Shaw authored
Not necessarily fatal(), but of potential interest when debugging odd slurmctld crashes. Cannot go where the limit is originally set, as the logging infrastructure is not avaiable at that point. Bug 3886.
-
Morris Jette authored
test1.91 fails with non-default binding
-
Tim Wickberg authored
-
Tim Shaw authored
Bug 3581.
-
Morris Jette authored
Fix for commit 250378c2 test7.3 was failing without this patch bug 3502
-
- 22 Jun, 2017 11 commits
-
-
Morris Jette authored
-
Morris Jette authored
test 17.12 was leaving slurm-#.out files around. Explicitly set output file to /dev/null and set time limit to 1 minute to avoid vestigial jobs.
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
Handle syncing between siblings (not the origin). Job could have been canclled or started while the origin was down. The sibling should see this and remove it's copy if the job is running or cancelled on the remote cluster. Handle case where job was cancelled or finished while the origin was down and the siblings were up.
-
Brian Christiansen authored
The controller will keeps job in the job_list until the origin comes back up and will find out about it then.
-