- 29 Oct, 2015 4 commits
-
-
Danny Auble authored
-
Danny Auble authored
submit_time. 0 means the job isn't eligible yet so we need to treat it so.
-
Danny Auble authored
-
Morris Jette authored
PrologFlags work: Setting a flag of "Contain" implicitly sets the "Alloc" flag. Fix code path which could prevent execution of the Prolog when the "Alloc" or "Contain" flag were set. bug 2074
-
- 28 Oct, 2015 2 commits
-
-
Morris Jette authored
If a user tries to create/destroy a persistent burst buffer and that is disabled (the default), return a permission denied error rather than an invalid request error. cray bug 832678
-
Tim Wickberg authored
Bug 2071
-
- 27 Oct, 2015 6 commits
-
-
Morris Jette authored
Log the minimum and maximum exit code (or signal) and show the state as "Mixed" if there are multiple termination states (e.g. cancelled, exit 0, exit 1, SIGBUS, etc.) bug 2058
-
Danny Auble authored
it needed to check for INFINITE64 instead of zero.
-
Danny Auble authored
-
Danny Auble authored
This just removes xfrees where the xfree had already happened.
-
Morris Jette authored
Correct scheduling logic for job features option with XOR construct that could delay a job's initiation. Some configurations and states with job feature XOR construct could prevent the job allocation from being made. bug 1865
-
Morris Jette authored
Improve sched/backfill support for job features, especially XOR construct. bug 1865
-
- 26 Oct, 2015 6 commits
-
-
Brian Christiansen authored
Bug 1973
-
Danny Auble authored
in the plugins.
-
Danny Auble authored
-
Danny Auble authored
when a user doesn't exist in the database.
-
Danny Auble authored
manager.
-
David Bigagli authored
-
- 24 Oct, 2015 2 commits
-
-
Morris Jette authored
Squeue recombines pending job array records only if their name and partition are identical. bug 2057
-
Morris Jette authored
Job array termination status email ExitCode based upon highest exit code from any task in the job array rather than the last task. bug 2058
-
- 23 Oct, 2015 2 commits
-
-
Hongjia Cao authored
Based upon patch from Hongjia Cao bug 2054
-
David Bigagli authored
-
- 22 Oct, 2015 4 commits
-
-
Morris Jette authored
-
Morris Jette authored
-
Brian Christiansen authored
-
Danny Auble authored
as it is the only TRES that can be given to another job while suspended.
-
- 21 Oct, 2015 2 commits
-
-
Morris Jette authored
sbatch --ntasks option to take precedence over --ntasks-per-node plus node count, as documented. Set SLURM_NTASKS/SLURM_NPROCS environment variables accordingly. bug 2015
-
David Bigagli authored
-
- 20 Oct, 2015 3 commits
-
-
Morris Jette authored
Avoid reporting more allocated CPUs than exist on a node. This can be triggered by resuming a previosly suspended job, resulting in oversubscription of CPUs. bug 2021
-
Danny Auble authored
-
Morris Jette authored
Add scancel -f/--full option to signal all steps including batch script and all of its child processes. bug 2031
-
- 19 Oct, 2015 6 commits
-
-
Brian Christiansen authored
Bug 1888
-
Danny Auble authored
out. Remove unneeded code that commit 8274ea54 fixed. This code would 0 out all GRES/TRES on a reconfig which isn't what we want. 8274ea54 does the right thing by itself.
-
Hongjia Cao authored
bug 2032
-
Morris Jette authored
Needed to change a couple of variables from 32- to 64-bit.
-
Morris Jette authored
Add new burst_buffer.conf parameters: ValidateTimeout and OtherTimeout. See man page for details.
-
David Bigagli authored
-
- 16 Oct, 2015 1 commit
-
-
David Bigagli authored
-
- 15 Oct, 2015 1 commit
-
-
Danny Auble authored
previously take 2 restarts of the slurmdbd to make it stick correctly.
-
- 14 Oct, 2015 1 commit
-
-
Danny Auble authored
single-threaded cores. A regression caused only 1 socket to be used on this kind of node instead of all that were available.
-