sh5utilSection: Slurm Commands (1)
Updated: Slurm Commands
sh5util - Tool for merging HDF5 files from the acct_gather_profile plugin that gathers detailed data for jobs running under Slurm
sh5util merges HDF5 files produced on each node for each step of a job into one HDF5 file for the job. The resulting file can be viewed and manipulated by common HDF5 tools such as HDF5View, h5dump, h5edit, or h5ls.
sh5util also has two extract modes. The first, writes a limited set of data for specific nodes, steps, and data series in "comma separated value" form to a file which can be imported into other analysis tools such as spreadsheets.
The second, (Item-Extract) extracts one data time from one time series for all the samples on all the nodes from a jobs HDF5 profile.
- - Finds sample with maximum value of the item.
- - Write CSV file with min, ave, max, and item totals for each node for each
- -E, --extract
Extract data series from a merged job file.
- Extract mode options
- -i, --input=path
- merged file to extract from (default ./job_$jobid.h5)
- -N, --node=nodename
- Node name to extract (default is all)
- -l, --level=[Node:Totals|Node:TimeSeries]
- Level to which series is attached. (default Node:Totals)
- -s, --series=[Energy | Lustre | Network | Tasks | Task_#]
Tasks is all tasks, Task_# (# is a task id) (default is everything)
- -I, --item-extract
Extract one data item from all samples of one data series from all nodes in a merged job file.
- -j, --jobs=<job(.step)>
- Format is <job(.step)>. Merge this job/step (or a comma-separated list of job steps). This option is required. Not specifying a step will result in all steps found to be processed.
- -h, --help
- Print this description of use.
- -o, --output=path
Path to a file into which to write. Default for merge is ./job_$jobid.h5 Default for extract is ./extract_$jobid.csv
- -p, --profiledir=dir
- Directory location where node-step files exist default is set in acct_gather.conf.
- -S, --savefiles
- Instead of removing node-step files after merging them into the job file, keep them around.
- User who profiled job. (Handy for root user, defaults to user running this command.)
Display brief usage message.
Data Items per Series
Reads Megabytes_Read Writes Megabytes_Write
Packets_In Megabytes_In Packets_Out Megabytes_Out
CPU_Frequency CPU_Time CPU_Utilization RSS VM_Size Pages Read_Megabytes Write_Megabytes
Executing sh5util sends a remote procedure call to slurmctld. If enough calls from sh5util or other Slurm client commands that send remote procedure calls to the slurmctld daemon come in at once, it can result in a degradation of performance of the slurmctld daemon, possibly resulting in a denial of service.
Do not run sh5util or other Slurm client commands that send remote procedure calls to slurmctld from loops in shell scripts or other programs. Ensure that programs limit calls to sh5util to the minimum necessary for the information you are trying to gather.
- Merge node-step files (as part of a sbatch script)
sbatch -n1 -d$SLURM_JOB_ID --wrap="sh5util --savefiles -j $SLURM_JOB_ID"
- Extract all task data from a node
sh5util -j 42 -N snowflake01 --level=Node:TimeSeries --series=Tasks
- Extract all energy data
sh5util -j 42 --series=Energy --data=power
COPYINGCopyright (C) 2013 Bull.
Copyright (C) 2013 SchedMD LLC. Slurm is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.
Slurm is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
This document was created by man2html using the manual pages.
Time: 22:00:05 GMT, April 09, 2020