- 12 Jun, 2013 4 commits
-
-
Janne Blomqvist authored
Signed-off-by:
Danny Auble <da@schedmd.com>
-
Janne Blomqvist authored
-
Morris Jette authored
-
Morris Jette authored
if on "scontrol reconfig" when AllowNodes is manually changed using scontrol since last slurmctld restart.
-
- 11 Jun, 2013 4 commits
-
-
Morris Jette authored
Conflicts: doc/man/man5/slurm.conf.5
-
Gennaro Oliva authored
-
Danny Auble authored
-
Danny Auble authored
correctly with poe when attempting to run > 32 nodes.
-
- 10 Jun, 2013 4 commits
-
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
due to either down nodes or explicit resizing. Generated slurmctld errors of this type: [2013-06-04T12:43:46+06:00] error: gres/gpu: step_test 68662.4294967294 gres_bit_alloc is NULL This is a movement of the logic introduced in commit https://github.com/SchedMD/slurm/commit/6fff97bb77d2d88aa808c47fd7880246a0c1d090 to eliminate a memory leak.
-
Morris Jette authored
bug 326
-
- 08 Jun, 2013 2 commits
-
-
Martin Perry authored
bug 327
-
Morris Jette authored
These changes partially address bug 326
-
- 07 Jun, 2013 7 commits
-
-
Rod Schultz authored
-
Rod Shultz authored
Our tester complained that srun --usage displays --profile=all rather than --profile=all|none|energy|lustre|network|task. None of the other options on usage have their possible options displayed so I made usage display --profile=... to be similar Then I enhanced --help to list the possible values. I also made the change to salloc and sbatch. bug 325
-
Danny Auble authored
overwritten by the parallel versions thus making it so we need handle both cases.
-
jette authored
-
Mark Nelson authored
we've been unable to submit jobs that request one task-per-core when we have SMT enabled (but it would work successfully for jobs that wanted two tasks-per-core). I ended up tracking it down to a regression caused by commit 8127554a. See bug 322
-
Morris Jette authored
-
Nathan Yee authored
-
- 06 Jun, 2013 5 commits
-
-
Rod Schultz authored
-
Morris Jette authored
-
Mark Nelson authored
-
Olli-Pekka Lehto authored
Thanks for the feedback! I was actually planning contributing a more complete set of documentation for the SLURM web pages but have been super busy. Feel free to copy the current documentation. I'll try to refine it a bit next week. It might be better to leave the link out as we might move the documentation to the "proper" CSC website at some point.
-
Morris Jette authored
-
- 05 Jun, 2013 14 commits
-
-
Danny Auble authored
-
Danny Auble authored
-
Danny Auble authored
-
Danny Auble authored
Since we don't currently track energy usage per task (only per step). Otherwise we get double the energy.
-
David Bigagli authored
-
Danny Auble authored
Conflicts: NEWS
-
Danny Auble authored
-
David Bigagli authored
-
Rod Schultz authored
-
Danny Auble authored
the user specifies a polling frequency larger than the default (Meaning the enforcement would happen at perhaps a much slower pace) then deny the job.
-
Danny Auble authored
-
Danny Auble authored
-
Morris Jette authored
-
Morris Jette authored
-