SLURM/ClusterStatus

From UMIACS
Revision as of 18:14, 19 December 2019 by Mbaney (talk | contribs)
Jump to navigation Jump to search

Cluster Status

SLURM offers a variety of tools to check the general status of nodes/partitions in a cluster.

sinfo

The sinfo command will show you the status of partitions in the cluster. Passing the -N flag will show each node individually.

tgray26@opensub00:sinfo
PARTITION AVAIL  TIMELIMIT  NODES  STATE NODELIST
dpart*       up   infinite      8   idle openlab[00-07]
gpu          up   infinite      2   idle openlab08
tgray26@opensub00:sinfo -N
NODELIST   NODES PARTITION STATE
openlab00      1    dpart* idle
openlab01      1    dpart* idle
openlab02      1    dpart* idle
openlab03      1    dpart* idle
openlab04      1    dpart* idle
openlab05      1    dpart* idle
openlab06      1    dpart* idle
openlab07      1    dpart* idle
openlab08      1       gpu idle

scontrol

The scontrol command can be used to view the status/configuration of the nodes in the cluster. If passed specific node name(s) only information about those node(s) will be displayed, otherwise all nodes will be listed. To specify multiple nodes, separate each node name by a comma (no spaces).

$ scontrol show nodes openlab00,openlab08
NodeName=openlab00 Arch=x86_64 CoresPerSocket=4
   CPUAlloc=8 CPUErr=0 CPUTot=8 CPULoad=7.10
   AvailableFeatures=(null)
   ActiveFeatures=(null)
   Gres=(null)
   NodeAddr=openlab00 NodeHostName=openlab00 Version=16.05
   OS=Linux RealMemory=7822 AllocMem=7822 FreeMem=149 Sockets=2 Boards=1
   State=ALLOCATED ThreadsPerCore=1 TmpDisk=49975 Weight=1 Owner=N/A MCS_label=N/A
   BootTime=2017-01-17T14:46:59 SlurmdStartTime=2017-01-17T14:47:43
   CapWatts=n/a
   CurrentWatts=0 LowestJoules=0 ConsumedJoules=0
   ExtSensorsJoules=n/s ExtSensorsWatts=0 ExtSensorsTemp=n/s


NodeName=openlab08 Arch=x86_64 CoresPerSocket=8
   CPUAlloc=1 CPUErr=0 CPUTot=16 CPULoad=1.19
   AvailableFeatures=(null)
   ActiveFeatures=(null)
   Gres=gpu:3
   NodeAddr=openlab08 NodeHostName=openlab08 Version=16.05
   OS=Linux RealMemory=128722 AllocMem=1024 FreeMem=395 Sockets=2 Boards=1
   State=MIXED ThreadsPerCore=1 TmpDisk=49975 Weight=1 Owner=N/A MCS_label=N/A
   BootTime=2016-12-22T20:26:52 SlurmdStartTime=2016-12-22T20:33:21
   CapWatts=n/a
   CurrentWatts=0 LowestJoules=0 ConsumedJoules=0
   ExtSensorsJoules=n/s ExtSensorsWatts=0 ExtSensorsTemp=n/s

sacctmgr

The sacctmgr command shows cluster accounting information. One of the helpful commands is to list the available QOS which translates into queues in systems like PBS/Torque.

$ sacctmgr list qos format=Name,Priority,MaxWall,MaxJobsPU
      Name   Priority     MaxWall MaxJobsPU
---------- ---------- ----------- ---------
    normal          0
     dpart          0  2-00:00:00         8
       gpu          0    08:00:00         2