Commits · 69b7ebd99482032a1400d4095633beae7faa1512 · Manuel G. Marciani / ces_slurm_simulator

21 Dec, 2018 7 commits
- run autogen.sh · 69b7ebd9
  Felip Moll authored Dec 11, 2018
```
(cherry picked from commit cc2a7ffe41b53fdc1fa4cf051fd0df2a6d982a31)
```
  69b7ebd9
- Initial commit for new xcc plugin. · efa71617
  Felip Moll authored Dec 11, 2018
```
This was copied directly from the ipmi plugin.  Only difference is
the acct_gather_energy_ipmi_config.* was combined into the main file.

Bug 5441

(cherry picked from commit 58681c0bd167d33596c2cd465d7cd502e1215439)
```
  efa71617
- run autogen.sh · 71c42419
  Danny Auble authored Dec 12, 2018
```
(cherry picked from commit 00f07df13b64b5bc82ce6a7d3de962793ae8af75)
```
  71c42419
- Add check for libfreeipmi needed for future xcc plugin · 78b4abc2
  Danny Auble authored Dec 12, 2018
```
instead of just libipmimonitoring (which already links to libfreeipmi).

Bug 5441

(cherry picked from commit 98d121a89ff59ee6c23262150050c3aac1714ded)
```
  78b4abc2
- Store ave watts in energy plugins. · a530e1ee
  Felip Moll authored Dec 21, 2018
```
Bug 5441
```
  a530e1ee
- Better debugging when a job doesn't have a job_resrcs ptr. · dc583bd1
  Dominik Bartkiewicz authored Dec 21, 2018
```
Bug 5971
```
  dc583bd1
- Don't assume the first node of a job is the batch host when purging jobs · 7da439b4
  Dominik Bartkiewicz authored Dec 21, 2018
```
from a node.

Bug 5971
```
  7da439b4
20 Dec, 2018 1 commit
- Read gres.conf for cloud nodes on slurmctld. · 1bceb1d9
  Brian Christiansen authored Dec 19, 2018
```
Bug 6229
```
  1bceb1d9
19 Dec, 2018 1 commit
- Fix double accounting of energy at end of job · 66e000a4
  Danny Auble authored Dec 19, 2018
```
Bug 6239
```
  66e000a4
18 Dec, 2018 5 commits
- Better debug for bad values in gres.conf · a1b0b2d9
  Moe Jette authored Dec 18, 2018
```
backport of 03ada72e

Bug 5682
```
  a1b0b2d9
- cons_res: Prevent overflow on multiply · df57ef17
  Dominik Bartkiewicz authored Feb 13, 2018
```
Bug 5682 and 4584

Backport of ba07a6e09b6071
```
  df57ef17
- Fix NodeFeaturesPlugins=node_features/knl_generic to allow other gres · 4d433d38
  Dominik Bartkiewicz authored Dec 18, 2018
```
other than knl.

Bug 5603
```
  4d433d38
- Insure that "hbm" is a configured GresType on knl systems. · c7d66c8d
  Danny Auble authored Dec 18, 2018
```
Bug 5603
```
  c7d66c8d
- Avoid bit offset of -1 in call to bit_nclear(). · 0e802d92
  Moe Jette authored Dec 18, 2018
```
This patch also simplifies the code using bit_cnt instead of
_bitstr_bits().  This isn't really necessary, but does look cleaner.

Bug 6216
```
  0e802d92
17 Dec, 2018 4 commits
- Update docs with actual reason message · db6a67e0
  Michael Hinton authored Dec 17, 2018
```
Bug 6420
```
  db6a67e0
- Check we are only dealing with one cluster when fixing runaways. · b1105048
  Danny Auble authored Dec 17, 2018
```
Continuation of 2 commits ago.

Bug 6159
```
  b1105048
- Sanity checks for empty lists when fixing runaway jobs. · 969591ae
  Broderick Gardner authored Dec 17, 2018
```
Continuation of last commit

Bug 6159
```
  969591ae
- Fix sacctmgr show runawayjobs from sibling cluster · f9adfede
  Broderick Gardner authored Dec 14, 2018
```
It was using the local cluster, sending the query to the wrong table

Bug 6159
```
  f9adfede
14 Dec, 2018 4 commits
- Set configurator to use info instead of numerical value for debug level. · c7a9d6a7
  Nate Rini authored Dec 12, 2018
```
bug 6116
```
  c7a9d6a7
- Set slurmdbd.conf man page example to use info instead of numerical value. · f58f537c
  Nate Rini authored Dec 12, 2018
```
bug 6116
```
  f58f537c
- Fix sacctmgr show events event=cluster · 1b0e01ba
  Brian Christiansen authored Dec 14, 2018
```
Bug 6237
```
  1b0e01ba
- Backfill - If a job has a time_limit guess the end time of a job better · 8c919742
  Dominik Bartkiewicz authored Dec 14, 2018
```
if OverTimeLimit is Unlimited.

Bug 6093
```
  8c919742
13 Dec, 2018 1 commit
- Run autogen.sh · d86853af
  Danny Auble authored Dec 13, 2018
  
  d86853af
11 Dec, 2018 3 commits
- Start NEWS for v18.08.5 · 17e96ba6
  Tim Wickberg authored Dec 11, 2018
  
  17e96ba6
- Update META for v18.08.4 release. · 35da90df
  Tim Wickberg authored Dec 11, 2018
```
Update slurm.spec and slurm.spec-legacy as well.
```
  35da90df
- Docs - change dual-factor to multi-factor. · 5a2e926b
  Tim Wickberg authored Dec 11, 2018
```
Bug 6029.
```
  5a2e926b
10 Dec, 2018 1 commit

Make CPU frequency test more forgiving · 8ee71c24

Morris Jette authored Dec 10, 2018

The cpu frequency set by the user is not exact with current kernels.
There seems to be a fair variation depending upon timing and other
events. This is resulting in test1.76 failing sporatically. This
changes the logic to retry if the frequency differs by more than
10 percent rather than failing immediately.

8ee71c24

08 Dec, 2018 1 commit
- pam_slurm_adopt html doc updates · 1a9a2149
  Marshall Garey authored Dec 07, 2018
```
Bug 6029
```
  1a9a2149
07 Dec, 2018 2 commits

salloc/sbatch/srun - print warning if both --mem and --mem-per-cpu are set. · 13a606a4

Nate Rini authored Dec 07, 2018

Only print a warning for 18.08. If a user has SLURM_MEM_PER_CPU or
SLURM_MEM_PER_NODE environment variables set for some reason this
situation could be happening by accident, and we don't want to prevent
the srun command from launching steps at this point.

Bug 6058.

13a606a4

Expand %x in 'scontrol show job' and related API calls. · 0a125e08
Broderick Gardner authored Dec 07, 2018
```
Bug 5648.
```
0a125e08

06 Dec, 2018 8 commits

Bump RLIMIT_NOFILE for daemons in systemd services · 7f2e6a7e

Janne Blomqvist authored Dec 05, 2018

The Linux kernel default hard limit of 4096 for the number of file
descriptors is quite small. Debian/Ubuntu have for a long time
overridden this, increasing it to 1M. Recently systemd also bumped the
default to 512k.

https://github.com/systemd/systemd/blob/master/NEWS

https://github.com/systemd/systemd/pull/10244

https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/thread/ZN5TK3D6L7SE46KGXICUKLKPX2LQISVX/

https://github.com/systemd/systemd/commit/09dad04c49cae3ad2b319c9b4e7773fedd34309a

Here the limits are increased as follows:

- slurmd: 128k; some workloads like Hadoop/Spark need a lot of fd's,
  and recommend that the limit is increased to at least 64k.

- slurmctld: 64k; per the Slurm high throughput and big system guides
  which recommend a file-max of at least 32k.

- slurmdbd: 64k, matching slurmctld, though slurmdbd shouldn't need
  that many fd's, bumping the limit shouldn't hurt either.

Bug 6171

7f2e6a7e

Fix formatting issues when printing uint64_t. · a04ae6d7
Tim Wickberg authored Dec 06, 2018
```
Bug 5248
```
a04ae6d7
job_submit/lua: Add user/group info to jobs. · 6bf90e85
Mike Nolta authored Dec 06, 2018
```
Bug 6055
```
6bf90e85

job_submit/lua: add several slurmctld return codes · c538e144

Mike Nolta authored Nov 16, 2018

Add the following slurmctld return codes to the lua plugin:

  ESLURM_ACCESS_DENIED
  ESLURM_ACCOUNTING_POLICY
  ESLURM_INVALID_NODE_COUNT
  ESLURM_JOB_MISSING_SIZE_SPECIFICATION
  ESLURM_MISSING_TIME_LIMIT

Bug 6055

c538e144

Docs - add note about requiring proctrack/cgroup when using PrologFlags=Contain. · 065d6554
Marshall Garey authored Dec 05, 2018
```
Bug 5836.
```
065d6554
Docs - add period at end of sentence. · d760a7ad
Marshall Garey authored Dec 05, 2018

d760a7ad
Fix Typo · 09e3a45e
Nate Rini authored Dec 05, 2018

09e3a45e
Fix typo · c9b5b295
Nate Rini authored Dec 05, 2018
```
Bug 6162
```
c9b5b295

05 Dec, 2018 2 commits
- Fix break lines for html version · b72b0c33
  Albert Gil authored Dec 05, 2018
```
Bug 6163
```
  b72b0c33
- Run SlurmctldPrimaryOffProg when the primary shuts down. · ee29bba8
  Felip Moll authored Dec 05, 2018
```
Backups already run it when dropping to backup.

Bug 6098.
```
  ee29bba8