- 11 Feb, 2016 1 commit
-
-
Morris Jette authored
Conflicts: src/slurmctld/read_config.c
-
- 10 Feb, 2016 20 commits
-
-
Morris Jette authored
chnage from 120 to 60 seconds. If it does not end in time, log it and orphan the process. For Cray systems with KNL, process boot times are likely to be ~20 minutes.
-
Morris Jette authored
-
Morris Jette authored
Eliminate duplicate code (put into function) and only power down node when needed.
-
Morris Jette authored
-
Brian Christiansen authored
-
Brian Christiansen authored
Bug 2427 Reproduced with: sacctmgr show user withassoc user= user=<username>
-
Danny Auble authored
-
Tim Wickberg authored
switch/federation superceeded by switch/nrt in a few places. Remove elan example runqsw.c entirely. Remove various checks from testsuite. Update documentation. HAVE_ELAN blocks should also go away.
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
Add new PowerParameters options of get_timeout and set_timeout. The default set_timeout was increased from 5 seconds to 30 seconds. Also re-read current power caps periodically or after any failed "set" operation. bug 2332
-
Tim Wickberg authored
Conflicts: src/scontrol/info_assoc_mgr.c
-
Tim Wickberg authored
Build should work on non-glibc distributions by better POSIX conformance.
-
Tim Wickberg authored
-
Tim Wickberg authored
Already had a mix of both standards, so no net change to build.
-
Tim Wickberg authored
Had already fixed this in FreeBSD build, so move that outside the #ifdef
-
Yu Watanabe authored
-
Brian Gilmer authored
This reverts commit 6c21c441. # Conflicts: # NEWS This is being reverted because we have been informed turning off NHC by default is not what Cray wants. We hope for a day when this will be the case and Slurm can run NHC on the compute node but until that day this is the way we will handle it.
-
Morris Jette authored
-
Morris Jette authored
-
- 09 Feb, 2016 10 commits
-
-
Morris Jette authored
This will let this program, spawned by slurmctld, log to SlurmctldLogFile (or some other configured file).
-
Tim Wickberg authored
Null implies jobacct_gather/none, and slurmd should not query slurmstepd for accounting information. Otherwise safe_read will block forever, and successive calls will max out the thread limit and slurmd will hang. Bug 2420.
-
Morris Jette authored
-
Morris Jette authored
-
Danny Auble authored
-
Danny Auble authored
-
Danny Auble authored
structure.
-
Morris Jette authored
-
Morris Jette authored
On a node with features that change with reboot and static features, preserve the static features on reboots
-
Alejandro Sanchez authored
-
- 08 Feb, 2016 6 commits
-
-
Morris Jette authored
If a job requires features which don't exist on all allocated nodes, then only reboot the nodes which lack the require features rather than all allocated nodes.
-
Danny Auble authored
-
Danny Auble authored
-
Danny Auble authored
which also makes it so we can use that instead of looking at CR_ONE_TASK_PER_CORE in the task plugin. This makes things layout correctly when using --ntasks-per-core which didn't work before.
-
Morris Jette authored
-
Morris Jette authored
-
- 06 Feb, 2016 2 commits
-
-
Tim Wickberg authored
-
Tim Wickberg authored
This is technically incorrect (although usually ignored), and will cause errors if compiling against slurm.h with -Wpedantic.
-
- 05 Feb, 2016 1 commit
-
-
Morris Jette authored
Get node name prefix and field width directly from actual node names.
-