sinfo

sinfo is used to list information about Slurm partitions and nodes in a cluster. It provides a quick overview of available resources, their states, and time limits.

Example:

CODE

[root@login02 ~]# sinfo 
PARTITION AVAIL TIMELIMIT NODES STATE NODELIST 
ai up infinite 14 mix ai[01-04,06,13,18-24,26] 
ai up infinite 1 alloc ai25 
ai up infinite 7 idle ai[05,07-08,11-12,14,16] 
avg up infinite 1 mix ai32 
avg up infinite 6 alloc ai[27-31,34] 
avg up infinite 6 idle ai[09-10,15,17,33,35] 
gm up infinite 1 mix gm01 
bdereli up infinite 1 idle comiecs01 
starlet up infinite 1 mix star03 
starlet up infinite 2 alloc star[01-02] 
starlet up infinite 3 idle star[04-06] 
kuttam up infinite 2 idle kuttam[01-02] 
cosbi up infinite 2 idle ag[01,04] 
kutem up infinite 1 idle rk01 
kutem_gpu up infinite 1 idle rk02 
lufer up infinite 1 idle it01 
short up 2:00:00 4 idle it[01-04] 
mid up 1-00:00:00 4 idle it[01-04] 
long up 7-00:00:00 4 idle it[01-04]

Command Output Columns

Column	Description
PARTITION	Partition (group) name
AVAIL	Partition availability status ( JAVA `up` / JAVA `down` )
TIMELIMIT	Maximum time limit for jobs in this partition
NODES	Number of nodes in this partition
STATE	Node state (resource usage and availability)
NODELIST	List of nodes in this partition

Node States and Their Meanings

State	Description
idle	Node is free — all resources are available
alloc	Node is fully allocated — no free resources
mix	Node is partially allocated — some resources are still free
down	Node is down or unreachable
draining	Node is being taken offline; not accepting new jobs, running jobs will finish first
drained	Node is completely drained and unavailable for any jobs