scontrol(1) Slurm Commands scontrol(1) (NAME) scontrol - Slurm. (SYNOPSIS) scontrol [OPTIONS...] [COMMAND...] (DESCRIPTION) scontrol Slurm : (job) (job step) (node) (partition) (reservation) . ( AUTHORIZATION ). . scontrol . . scontrol . (case-insensitive) ( "LX" "lx" ). . Slurm scontrol write config . "slurm.conf." "slurm.conf" . slurm.conf . (OPTIONS) -a, --all show . . -M, --clusters= . . slurmdbd FederationParameters=fed_display --federation . --local . -d, --details show . --federation . -F, --future FUTURE . -h, --help scontrol . --hide . ( ). --json, --json=list, --json= JSON data_parser data_parser . . . . --details . --local . . --federation . -o, --oneliner . -Q, --quiet . --sibling (sibling jobs) . --federation . -u, --uid= . -v, --verbose . '-v' . . -V , --version . --yaml, --yaml=list, --yaml= YAML data_parser data_parser . . . . --details . cancel_reboot . ASAP (drained) (undrain) . create . . completing COMPLETING COMPLETING DOWN . delete SPECIFICATION . SPECIFICATION : NodeName= . . ( "lx[10-20]"). (Nodesets) . "ALL" ( ). PartitionName= . . . slurm.conf . ReservationName= . . . EndTime Duration . . SlurmctldParameters=user_resv_delete slurm.conf Flags=USER_DELETE . errnumstr Slurm . fsdampeningfactor FairShareDampeningFactor slurmctld. getaddrs IP slurmctld. getent [] ( ) . slurmstepd slurmctld . --enable-multiple-slurmd . : LaunchParameters=disable_send_gids slurm.conf . help scontrol. hold ( ). job_list ( SLUID) "jobname=" . . admin-hold user-hold . release ( uhold ). . listjobs [] . slurmstepd slurmctld . --enable-multiple-slurmd . listpids [[.]] [] (PID) ( JOBID.STEPID ) ( job_id ) ( job_id job_id "*" ). scontrol Slurm . Slurm ( pgid ProctrackType) . NodeName slurmd . slurmd Slurm . liststeps [] . slurmstepd slurmctld . --enable-multiple-slurmd . notify salloc srun (batch job) job_id . pidinfo Slurm proc_id . scontrol Slurm . ping slurmctld . --clusters slurmctld . power {up|down} [asap|force] {ALL||} [Action=] [Reason=] (power) / . 'power down' ASAP power up . / . reason (power down) . "action" PowerAction ResumeProgram SuspendProgram . power up Location=slurmctld . update scontrol : scontrol update NodeName= State={POWER_UP|POWER_DOWN|POWER_DOWN_ASAP|POWER_DOWN_FORCE} : down SuspendProgram () . SuspendTimeout . ResumeProgram capmc Cray "IDLE" . "power down" "power up" "IDLE" . down asap () (drain) . () . down force () "IDLE" ..IP up ResumeProgram () (power saving) . ResumeTimeout . reboot [ASAP] [FORCE] [nextstate={RESUME|DOWN}] [reason=][action=] {ALL||} (idle) RebootProgram slurm.conf . "action" PowerAction . "REBOOT" . slurmd HealthCheckProgram . slurmd slurmctld "REBOOT" . "ASAP" "DRAIN" << >> (ASAP) . "reason" "ASAP" "Reboot ASAP" nextstate nextstate=UNDRAIN . "FORCE" reboot power (requeue) . "nextstate" "DOWN" down . "nextstate" "RESUME" (reason) "DRAIN" . backfill . "reason" . "reboot requested" . : "reboot issued" "reboot complete" "nextstate" "DOWN" "reboot timed out" ResumeTimeout . (ALL) . : : FUTURE POWER_DOWN POWERED_DOWN POWERING_DOWN REBOOT_ISSUED REBOOT_REQUESTED : . "ASAP" drain . (advanced reservation) . "scontrol cancel_reboot " "CANCEL_REBOOT" . ResumeTimeout "DOWN" . reconfigure slurmctld slurmd . slurm.conf(5) . 23.11 . . TCP . release (held) . job_list ( SLUID) "jobname=" . admin-hold user-hold . user-hold . hold. requeue [