Slurm: Difference between revisions

From Halfface
Jump to navigation Jump to search
(Created page with "==install slurm under fedora 21== # Build slurm rpmbuild -ta slurm*.tar.bz2 # Install rpms. yum -y install munge slurm slurm-plugins slurm-munge # configure munge dd if=/dev/...")
 
 
(7 intermediate revisions by the same user not shown)
Line 1: Line 1:
==install slurm under fedora 21==
==install slurm under fedora 26==
# Build slurm
Edit slurm.spec to avoid: "plugin_load_from_file: dlopen(/usr/lib64/slurm/select_linear.so): /usr/lib64/slurm/select_linear.so: undefined symbol: slurm_job_preempt_mode." Add the four lines between %...
%build
CFLAGS="$RPM_OPT_FLAGS -Wl,-z,lazy"
CXXFLAGS="$RPM_OPT_FLAGS -Wl,-z,lazy"
export CFLAGS
export CXXFLAGS
%configure \
 
Build slurm
  rpmbuild -ta slurm*.tar.bz2
  rpmbuild -ta slurm*.tar.bz2
# Install rpms.
Install rpms.
yum -y install munge slurm slurm-plugins slurm-munge
yum -y install munge slurm slurm-plugins slurm-munge
# configure munge
 
dd if=/dev/random bs=1 count=1024 > /etc/munge/munge.key
==configure munge==
chmod 0600 /etc/munge/munge.key
dd if=/dev/random bs=1 count=1024 > /etc/munge/munge.key
chown munge /etc/munge/munge.key
chmod 0600 /etc/munge/munge.key
systemctl start munge
chown munge /etc/munge/munge.key
# Change config to better suit fedora 21
systemctl start munge
SlurmctldPidFile=/var/run/slurm/slurmctld.pid
 
SlurmdPidFile=/var/run/slurm/slurmd.pid
StateSaveLocation=/var/lib/slurm/slurm.state
SlurmdSpoolDir=/var/lib/slurm/slurm.spool
# create corresponding directories.
DIR=/var/run/slurm;chown slurm:slurm $DIR ; chmod 755 $DIR
# update systemd config files for slurm daemons to point to new location of pid file.
vim /usr/lib/systemd/system/slurmctld.service /usr/lib/systemd/system/slurmd.service
==test installation==
==test installation==
# Test the installation.
Generate a credential on stdout.
Generate a credential on stdout.
  munge -n
  munge -n
Line 28: Line 28:
Run a quick benchmark.
Run a quick benchmark.
  remunge
  remunge
# how does it work
how does it work
  scontrol show config
  scontrol show config
# check priorities of jobs using the command
check priorities of jobs using the command
  scontrol show job".
  scontrol show job".
# Submit a job
==job control==
Submit a job
  sbatch /tmp/slurm_test_1
  sbatch /tmp/slurm_test_1
# List jobs:
List jobs:
  squeue
  squeue
# Get job details:
Get job details:
  scontrol show job 106
  scontrol show job 106
# Suspend a job (root only):
Suspend a job (root only):
  scontrol suspend 135
  scontrol suspend 135
# Resume a job (root only):
Resume a job (root only):
  scontrol resume 135
  scontrol resume 135
# Kill a job. Users can kill their own jobs, root can kill any job.
Kill a job. Users can kill their own jobs, root can kill any job.
  scancel 135
  scancel 135
# Hold a job
Hold a job
  scontrol hold 139
  scontrol hold 139
# Release a job:
Release a job:
  scontrol release 139
  scontrol release 139
# List partitions:
List partitions:
  sinfo
  sinfo
# example job script.
example job script.
  #!/usr/bin/env bash
  #!/usr/bin/env bash
  #SBATCH -p defq
  #SBATCH -p defq
  #SBATCH -J simple
  #SBATCH -J simple
  sleep 60
  sleep 60
==troubleshooting==
Restore node.
scontrol update nodename=www state=down Reason=troubleshooting
scontrol update nodename=www state=resume

Latest revision as of 17:05, 5 November 2017

install slurm under fedora 26

Edit slurm.spec to avoid: "plugin_load_from_file: dlopen(/usr/lib64/slurm/select_linear.so): /usr/lib64/slurm/select_linear.so: undefined symbol: slurm_job_preempt_mode." Add the four lines between %... 

%build
CFLAGS="$RPM_OPT_FLAGS -Wl,-z,lazy"
CXXFLAGS="$RPM_OPT_FLAGS -Wl,-z,lazy"
export CFLAGS
export CXXFLAGS
%configure \

Build slurm 

rpmbuild -ta slurm*.tar.bz2

Install rpms. 

yum -y install munge slurm slurm-plugins slurm-munge

configure munge

dd if=/dev/random bs=1 count=1024 > /etc/munge/munge.key
chmod 0600 /etc/munge/munge.key
chown munge /etc/munge/munge.key
systemctl start munge

test installation

Generate a credential on stdout. 

munge -n

Check if a credential can be locally decoded. 

munge -n | unmunge

Check if a credential can be remotely decoded. 

munge -n | ssh somehost unmunge

Run a quick benchmark. 

remunge

how does it work 

scontrol show config

check priorities of jobs using the command 

scontrol show job".

job control

Submit a job 

sbatch /tmp/slurm_test_1

List jobs: 

squeue

Get job details: 

scontrol show job 106

Suspend a job (root only): 

scontrol suspend 135

Resume a job (root only): 

scontrol resume 135

Kill a job. Users can kill their own jobs, root can kill any job. 

scancel 135

Hold a job 

scontrol hold 139

Release a job: 

scontrol release 139

List partitions: 

sinfo

example job script. 

#!/usr/bin/env bash
#SBATCH -p defq
#SBATCH -J simple
sleep 60

troubleshooting

Restore node. 

scontrol update nodename=www state=down Reason=troubleshooting
scontrol update nodename=www state=resume
Retrieved from "https://halfface.se/wiki/index.php?title=Slurm&oldid=12915"