Commits · 714cd570631afa41fde539297a5bda920a88f8ff · tud-zih-energy / Slurm

Mar 07, 2014
- NRT - change xmalloc's to malloc just to be safe · 714cd570
  Danny Auble authored 11 years ago
  
  714cd570
- NRT - Fix issue where we are launching hosts out of numerical order, · d775d2b3
  Danny Auble authored 11 years ago
  
  this would cause pmd's to hang.
  d775d2b3
- NRT - initialize missing variables when the PMD is starting a job · 7b7c6006
  Danny Auble authored 11 years ago
  
  7b7c6006
- NRT - Increase Max number of adapters from 8 -> 9 · 7015345e
  Danny Auble authored 11 years ago
  
  7015345e
- NRT - Fix issue where there are more than NRT_MAXADAPTERS on a system. · 6d2b6cf2
  Danny Auble authored 11 years ago
  
  6d2b6cf2
Mar 06, 2014
- NRT - Fix possible memory leak if using multiple adapters. · dfb66218
  Danny Auble authored 11 years ago
  
  dfb66218
- Remove need to use job->ctx_params in the launch plugin, just to simplify · 0f6cce43
  Danny Auble authored 11 years ago
  
  code.
  0f6cce43
- Switch/nrt: Correct get dynamic window count logic · 374bd61e
  Morris Jette authored 11 years ago
  
  374bd61e
Mar 04, 2014
- Fix the database query to return all pending jobs in a given time · ea7d6078
  Bill Brophy authored 11 years ago
  
  interval.
  ea7d6078
- Add munge connect retry · cece852a
  Morris Jette authored 11 years ago
  
  cece852a
- Make S_SLURM_RESTART_COUNT item available to SPANK · 96673989
  Magnus Jonsson authored 11 years ago
  
  96673989
- Avoid crash if gres configuration inconsistent · f005e508
  Carlos Bederián authored 11 years ago
  
  f005e508
Mar 03, 2014
- Fix sinfo output so that host in state allocated and mixed will not be · 8cff3b08
  David Bigagli authored 11 years ago
  
  merged together.
  8cff3b08
- Don't print negative number when dealing with large memory sizes with · 82d48a0d
  Josko Plazonic authored 11 years ago
  
  sacct.
  82d48a0d
Mar 02, 2014

Add SchedulerParameters option bf_max_job_start · 1b0c4a33

jette authored 11 years ago

Add support for SchedulerParameters value of bf_max_job_start that limits
the total number of jobs that can be started in a single iteration of the
backfill scheduler.
bug 607

1b0c4a33

Feb 27, 2014
- Add SLURM_STEP_ID to Prolog environment · 34da34fd
  Morris Jette authored 11 years ago
  
  bug 607
  34da34fd
- NRT - Fix to supply correct error messages to poe/pmd when a launch fails. · b0460a3d
  Danny Auble authored 11 years ago
  
  b0460a3d
- MYSQL - Fix when updating QOS on an association. · 324a089e
  Danny Auble authored 11 years ago
  
  324a089e
Feb 26, 2014
- MYSQL - Fixed memory leak when querying clusters · c63eb7e6
  Danny Auble authored 11 years ago
  
  c63eb7e6
- Fixed minor memory leak in backfill scheduler. · 911beb99
  Danny Auble authored 11 years ago
  
  911beb99
Feb 25, 2014
- BGQ - Fix minor memory leak when selecting blocks that can't immediately be · 1330681b
  Danny Auble authored 11 years ago
  
  placed.
  1330681b
- Update sacct man page description of job states. · 57b3f0c4
  David Bigagli authored 11 years ago
  
  57b3f0c4
Feb 21, 2014
- Fix minor memory leak when adding a reservation with a nodelist and core · 950231cc
  Danny Auble authored 11 years ago
  
  count.
  950231cc
- Fix minor memory leak when updating a reservation on a partition using "ALL" · d0ab7ba8
  Danny Auble authored 11 years ago
  
  nodes.
  d0ab7ba8
- Fix minor memory leak when updating a job's name. · 7211ce17
  Danny Auble authored 11 years ago
  
  7211ce17
Feb 20, 2014

Scheduling fix for jobs with specific nodes required · eafc0a4f

Morris Jette authored 11 years ago

If a job requires specific nodes and can not run due to those nodes being
busy, the main scheduling loop will block those specific nodes rather than
the entire queue/partition.
bug 595

eafc0a4f

Fix typo in NEWS · 5db06c04
Morris Jette authored 11 years ago

5db06c04

Feb 19, 2014
- Fix slurmctld core dump when a jobs gets its qos updated but there · 2e0f8062
  David Bigagli authored 11 years ago
  
  is not a corresponding association.
  2e0f8062
- Fix slurmctld core dump when a jobs gets its qos updated but there · 72a53d4f
  David Bigagli authored 11 years ago
  
  is not a corresponding association.
  72a53d4f
Feb 14, 2014
- Update srun.1 man page documenting the PMI2 support. · 2c83c36e
  David Bigagli authored 11 years ago
  
  2c83c36e
- Fix issue where if using munge and munge wasn't running and a slurmd · ddc0b5c3
  Danny Auble authored 11 years ago
  
  needed to forward a message the slurmd would core dump.
  ddc0b5c3
Feb 13, 2014
- Correct the slurm.conf man pages and checkpoint_blcr.html page · b5a79c9f
  David Bigagli authored 11 years ago
  
  describing that jobs must be drained from cluster before deploying any checkpoint plugin.
  b5a79c9f
Feb 12, 2014

enforce cpus-per-task with mem-per-cpu option · cf367bb0

Morris Jette authored 11 years ago

Properly enforce a job's cpus-per-task option when a job's allocation is
constrained on some nodes by the mem-per-cpu option.
bug 590

cf367bb0

Feb 10, 2014
- Updates for start of v2.6.7 work · 2c9b35c3
  Morris Jette authored 11 years ago
  
  2c9b35c3
Feb 09, 2014
- CRAY - fix memory leak when using accelerators · ac64f883
  Moe Jette authored 11 years ago
  
  ac64f883
Feb 08, 2014
- replace old commit note erroneously taken out · 099f00ab
  Danny Auble authored 11 years ago
  
  099f00ab
- CRAY - fix issue with using CR_ONE_TASK_PER_CORE · fbb37db1
  Danny Auble authored 11 years ago
  
  fbb37db1
Feb 07, 2014
- Properly enforce GrpSubmit limit for job arrays. · 9469053d
  Morris Jette authored 11 years ago
  
  bug 586
  9469053d
Feb 05, 2014
- take back 2.6.6 · 9f97c2e9
  Danny Auble authored 11 years ago
  
  9f97c2e9
- Added support for selecting AMD GP · f728ee8e
  Dominik Bartkiewicz authored 11 years ago
  
  Set GPU_DEVICE_ORDINAL environment variable.
  f728ee8e