- 06 Jan, 2016 2 commits
-
-
Tim Wickberg authored
salloc/sbatch/srun did not mention this. Also reference OverTimeLimit as another option affecting the final run time. Bug 2309.
-
Danny Auble authored
the job starts update the cpus_per_task appropriately. This also moves update num_tasks to after the setting of node counts on an update. It didn't appear to matter, but the cpus_per_task and pn_min_cpus had to be figured out after the cpus and nodes were set but before tasks. Bug 2302
-
- 05 Jan, 2016 3 commits
-
-
Morris Jette authored
burst_buffer/cray - Improve tracking of allocated resources to handle race condition when reading state while buffer allocation is in progress. Also initialize a mutex
-
Danny Auble authored
DBD for the first time. The corruption is only noticed at shutdown. Bug 2293
-
Morris Jette authored
-
- 04 Jan, 2016 4 commits
-
-
Morris Jette authored
Set job's reason to "Priority" when higher priority job in that partition (or reservation) can not start rather than leaving the reason set to "Resources". bug 2285
-
Danny Auble authored
error message.
-
Danny Auble authored
-
Morris Jette authored
If a reservation's nodes value is "all" then track the current nodes in the system, even if those nodes change. Nodes will automatically be added to or removed from a reservation when slurm.conf changes. bug 2204
-
- 02 Jan, 2016 1 commit
-
-
Brian Christiansen authored
Bug 2281
-
- 31 Dec, 2015 2 commits
-
-
Tim Wickberg authored
Rename the variable to match rest of codebase while here. This is related to bug 2295, although snprintf() protects against buffer overflow in 15.08 and up.
-
Tim Wickberg authored
Later releases have switched over to snprintf to avoid this issue, but 14.11 did not get that patch. Bug 2295.
-
- 30 Dec, 2015 1 commit
-
-
Danny Auble authored
-
- 29 Dec, 2015 3 commits
-
-
Alejandro Sanchez authored
time.
-
Danny Auble authored
static/overlap systems when some hardware issue happens when restarting the slurmctld.
-
Danny Auble authored
a dynamic system and mark the block in error on a static/overlap system. Bug 2273
-
- 28 Dec, 2015 2 commits
-
-
Morris Jette authored
Don't use lower weight nodes for job allocation when topology/tree used. bug 2284
-
Morris Jette authored
Preemption/gang scheduling: If a job is suspended at slurmctld restart or reconfiguration time, then leave it suspended rather than resume+suspend. bug 2274
-
- 23 Dec, 2015 2 commits
-
-
Morris Jette authored
task/affinity: Disable core-level task binding if more CPUs required than available cores. bug 2267
-
Morris Jette authored
Log as error if more than 3 aeld connects per second that cause is likely duplicate slurmctld daemon bug 2278
-
- 22 Dec, 2015 1 commit
-
-
Morris Jette authored
This is needed to properly enforce limits and account for usage.
-
- 19 Dec, 2015 1 commit
-
-
John Hensley authored
Remove the 1024-character limit on lines in batch scripts, which was causing long lines to be silently truncated. I noticed it when jobs were getting created with fewer dependencies than specified. Also increase the line length when showing job info.
-
- 18 Dec, 2015 3 commits
-
-
jette authored
sched/backfill: If a job can not be started within the configured backfill_window, set it's start time to 0 (unknown) rather than the end of the backfill_window. bug 2100
-
Danny Auble authored
-
jette authored
If a pending job array has multiple reasons for being in a pending state, then print all reasons in a comma separated list. Before: JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) 97354_[1-4] debug tmp jette PD 0:00 1 (Resources) After: JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) 97354_[1-4] debug tmp jette PD 0:00 1 (Resources,JobHeldUser)
-
- 16 Dec, 2015 6 commits
-
-
Brian Christiansen authored
Bug 2130
-
Morris Jette authored
Move slurmctld mail handler to separate thread for improved performance. Original logic did fork/exec without separate thread and if the slurmctld memory size is huge, then the time required for fork() to complete can be significant. bug 2252
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
Avoid hard-link/copy of script/environment files for job arrays. Use the master job record file for all tasks of the job array. back port of commit 41eead64 bug 2220
-
Morris Jette authored
Refactor logic to synchronize active bach jobs and their script/environment files, reducing overhead dramatically for large numbers of active jobs. backport of commit 41eead64 bug 2220
-
- 15 Dec, 2015 6 commits
-
-
Morris Jette authored
If a job can never be started due to unsatisfied job dependencies, report the full original job dependency specification rather than the dependencies remaining to be satisfied (typically NULL). bug 2197
-
Danny Auble authored
in use.
-
Danny Auble authored
_slurm_rpc_complete_job_allocation. This is a rewrite of 438365ec which didn't catch the job_ptr wasn't in a lock so the memory issue could still of existed. This hopefully fixes all the spots the job_ptr wasn't in the lock. Fixes bug 2146
-
Brian Christiansen authored
Bug 2261
-
Danny Auble authored
15.08.5. Bug 2243
-
David Bigagli authored
bug 2171
-
- 14 Dec, 2015 2 commits
-
-
Morris Jette authored
Decrease parallelism in job cancel request to prevent denial of service when cancelling huge numbers of jobs. bug 2256
-
Morris Jette authored
Prevent triggering gang scheduling within a partition if configured with PreemptType=partition_prio and PreemptMode=suspend,gang. The essence of this fix is to change a "<=" to "<" in cons_res/job_test.c: - if ((p_ptr->part_ptr->priority <= jp_ptr->part_ptr->priority) && + if ((p_ptr->part_ptr->priority < jp_ptr->part_ptr->priority) && but logic was also added to insure that a partition configuration with PreemptMode did not override PreemptType != partition_prio. bug 2232
-
- 11 Dec, 2015 1 commit
-
-
Tim Wickberg authored
Previously an error() would be logged when the attempt to open the job script using the new directory format failed but the successive fallback to the old directory structure was successful, leading to confusion when troubleshooting. Move emitted warnings to debug(), and only error() after failing to open in both directory structures. Add a note about backwards compatibility to both functions - we cannot remove these fallbacks as directory structure for pending jobs does not change on Slurm version update, and people may need to chain multiple version update together to get to a current slurm version which would correctly update slurmctld state files but leave pending jobs in the old directory structure. Bug #2244.
-