Code

Add Slurm support#42

Code

Open

aaronk wants to merge 27 commits intobenedictpaten:masterbenedictpaten/jobTree:masterfrom harvardinformatics:masterharvardinformatics/jobTree:masterCopy head branch name to clipboard

+4,452-42

Commits

Commits on Jan 9, 2015

• slurm additions - doesn't work

  

  Aaron Kitzmiller

  committed
  b29ce56

  Copy full SHA for b29ce56

Commits on Jan 16, 2015

• mostly functional slurm batch system

  

  Aaron Kitzmiller

  committed
  164839d

  Copy full SHA for 164839d

Commits on Jan 20, 2015

• turned off random errors in sort test; added JT_SLURM envs for passing slurm partition and time

  

  Aaron Kitzmiller

  committed
  d52838b

  Copy full SHA for d52838b

• actually doing it correctly

  

  Aaron Kitzmiller

  committed
  0c01cb0

  Copy full SHA for 0c01cb0

Commits on Jan 22, 2015

• dynamic batch system loading and command line option definition

  

  Aaron Kitzmiller

  committed
  cfdb615

  Copy full SHA for cfdb615

• slurmopts successfully being passed to SlurmBatchSystem and Worker; batchSystemConstructionFn reconstituted with generic code so that bigBatchSystem can be constructed like others; some documentati…

  Show description for 634e9df
  

  Aaron Kitzmiller

  committed
  634e9df

  Copy full SHA for 634e9df

• Adding 14.03.8 as default

  ifxdeploy

  committed
  8ffe9bf

  Copy full SHA for 8ffe9bf

Commits on Jan 23, 2015

• Converted default --slurm-time to string.

  ifxdeploy

  committed
  273e916

  Copy full SHA for 273e916

• Really actually fix the default slurm time

  ifxdeploy

  committed
  3a98c34

  Copy full SHA for 3a98c34

Commits on Mar 5, 2015

• Added multiple attempts for sbatch in case of failure. Sleep time between attempts increases via fibonacci until it gets to about 1.5 hours

  ifxdeploy

  committed
  25754fd

  Copy full SHA for 25754fd

Commits on Mar 11, 2015

• Modify processAnyNewFile to treat presence of a .new file as failure of atomic update.

  tsackton

  committed
  b0f4c4a

  Copy full SHA for b0f4c4a

• added print out and retries for zero length pickle files

  ifxdeploy

  committed
  0da218b

  Copy full SHA for 0da218b

• Modify processAnyNewFile to treat presence of a .new file as failure of atomic update.

  tsackton

  committed
  4bbd380

  Copy full SHA for 4bbd380

Commits on Mar 16, 2015

• Adding retry code to sacct and scancel methods

  tsackton

  committed
  d9f7dbe

  Copy full SHA for d9f7dbe

• Adding retry code to sacct and scancel methods

  tsackton

  committed
  9484916

  Copy full SHA for 9484916

Commits on Apr 1, 2015

• capture exception; return none

  ifxdeploy

  committed
  62fddf6

  Copy full SHA for 62fddf6

• Merge branch 'master' of github.com:harvardinformatics/jobTree

  ifxdeploy

  committed
  f1131f8

  Copy full SHA for f1131f8

Commits on Apr 17, 2015

• Double memory after job failure (I hope).

  tsackton

  committed
  775f4fc

  Copy full SHA for 775f4fc

Commits on Jun 24, 2015

• Added COMPLETING to the finished states and make /usr/bin/squeue the binary for squeue

  ifxdeploy

  committed
  b307fc0

  Copy full SHA for b307fc0

Commits on Jun 28, 2015

• Attempting to fix memory doubling after failed job.

  tsackton

  committed
  f5248b5

  Copy full SHA for f5248b5

Commits on Feb 23, 2016

• support for job name pattern

  fasifx

  committed
  72bf060

  Copy full SHA for 72bf060

• remove quotes from --jobname pattern

  fasifx

  committed
  ebac036

  Copy full SHA for ebac036

Commits on Mar 22, 2016

• /usr/local/bin/squeue

  aaronk

  committed
  a80e92f

  Copy full SHA for a80e92f

Commits on Oct 18, 2016

• added --constraint

  

  =

  committed
  c17eac4

  Copy full SHA for c17eac4

• added --constraint

  

  =

  committed
  59e71fe

  Copy full SHA for 59e71fe

Commits on Mar 20, 2017

• adding qos

  

  =

  committed
  50e6f53

  Copy full SHA for 50e6f53

• adding qos

  

  =

  committed
  b4bdaeb

  Copy full SHA for b4bdaeb