Overview

Installed using apt (see below), config based on
- https://slurm.schedmd.com/quickstart_admin.html
- use configurator: https://slurm.schedmd.com/configurator.html
- results pasted into /etc/slurm-llnl/slurm.conf
- Note for standalone controller and compute node I used:
- NodeName=deq2
- SlurmctldHost=deq2
Note: slurm prior to 20.11.7 or 20.02.7 are no longer available due to a security issue (Ubuntu dpkg is 19.x, has this been patched)
install slurm with apt
This blog has some possibly good info for a standalone cluster on 20.04:
- https://blog.llandsmeer.com/tech/2020/03/02/slurm-single-instance.html
- https://gist.github.com/ckandoth/2acef6310041244a690e4c08d2610423

slurm.conf options

/etc/

Showing only uncommented lines below (with cat /etc/slurm-llnl/slurm.conf |grep -v "#")


ClusterName=vahydro
SlurmctldHost=deq2
MpiDefault=none
ProctrackType=proctrack/cgroup
ReturnToService=1
SlurmctldPidFile=/var/run/slurmctld.pid
SlurmctldPort=6817
SlurmdPidFile=/var/run/slurmd.pid
SlurmdPort=6818
SlurmdSpoolDir=/var/spool/slurmd
SlurmUser=slurm
StateSaveLocation=/var/spool/slurmctld
SwitchType=switch/none
TaskPlugin=task/affinity
InactiveLimit=0
KillWait=30
MinJobAge=300
SlurmctldTimeout=120
SlurmdTimeout=300
Waittime=0
SchedulerType=sched/backfill
SelectType=select/cons_tres
SelectTypeParameters=CR_Core
AccountingStorageType=accounting_storage/none
JobCompLoc=/opt/model/slurm/jobs.log
JobCompType=jobcomp/filetxt
JobAcctGatherFrequency=30
JobAcctGatherType=jobacct_gather/none
SlurmctldDebug=info
SlurmctldLogFile=/var/log/slurmctld.log
SlurmdDebug=info
SlurmdLogFile=/var/log/slurmd.log
NodeName=deq2 CPUs=1 CoresPerSocket=8 ThreadsPerCore=2 State=UNKNOWN
PartitionName=debug Nodes=ALL Default=YES MaxTime=INFINITE State=UP


#### apt install slurm 
- old version

sudo apt install slurm-wlm mkdir /opt/model/slurm sudo chown slurm:modelers /opt/model/slurm sudo mkdir /var/spool/slurmctld sudo chown slurm:modelers /var/spool/slurmctld

configurator.html

nano /etc/slurm-llnl/slurm.conf

copy basic cgroup.conf file

cp /usr/share/doc/slurmd/examples/cgroup.conf /etc/slurm-llnl/

get it all ready to run automatically

sudo systemctl enable munge sudo systemctl enable slurmctld sudo systemctl enable slurmd

start them up

sudo systemctl start munge sudo systemctl start slurmctld sudo systemctl start slurmd

#### apt install slurmdbd
- `sudo apt install slurmdbd`

##### Configure slurmdbd

- The install above seems to work, `sinfo` appears to generate reasonable info:

sinfo PARTITION AVAIL TIMELIMIT NODES STATE NODELIST debug up infinite 1 unk deq2



- Make it go online
  - `sudo scontrol update NodeName=deq2 State=resume`
- list active jobs
  - `squeue`
- cancel a job
  - `scancel 12`

Had problem starting slurmd (slurmctld started OK if slurm.conf is OK) Needed to get cgroup.conf set up cp /usr/share/doc/slurmd/examples/cgroup.conf /etc/slurm-llnl/

Problem: slurmd wil not start on deq2 , so jobs never run on : running sudo systemctl status slurmd.service yields:


● slurmd.service - Slurm node daemon
     Loaded: loaded (/lib/systemd/system/slurmd.service; enabled; vendor preset: enabled)
     Active: failed (Result: exit-code) since Wed 2022-02-23 19:03:00 UTC; 2min 1s ago
       Docs: man:slurmd(8)
    Process: 512228 ExecStart=/usr/sbin/slurmd $SLURMD_OPTIONS (code=exited, status=1/FAILURE)

Feb 23 19:03:00 deq2 systemd[1]: Starting Slurm node daemon...
Feb 23 19:03:00 deq2 systemd[1]: slurmd.service: Control process exited, code=exited, status=1/FAILURE
Feb 23 19:03:00 deq2 systemd[1]: slurmd.service: Failed with result 'exit-code'.
Feb 23 19:03:00 deq2 systemd[1]: Failed to start Slurm node daemon.

HARPgroup / cbp_wsm