nodes of the cluster.

and slurmctld (specify different configuration file).

SLURM credentials.

be expected.

than errno (which is not reset by the function).

MAX_SERVER_THREADS and increase that value from 50 to 60

results from apparent NFS synchronization delays).

name a prefix of "slurm_".

flag the required state saves. Perform the saves using synchronous I/O
from just one thread. Under heavy loads this results in much faster
responsiveness and lowers slurmctld's memory and CPU overhead considerably.

 o also show node state when -l is used with -R
 o Don't let environment var override command line (SINFO_FORMAT)

  of reasons and nodes with those reasons.

right away and treat failure as a fatal error.

write to.

    slurmctld startup
Create StateSaveLocation directory if changes via slurmctld reconfig

Properly support sinfo --state option value including "*" suffix.

  (which exhaust available memory).
Restructure slurmctld agent logic for higher throughput.

Fixes problem running very large number of simultaneous jobs when
their inactivity limit was reached due to a backlog of message.

data structure. This eliminates risks associated with re-reading slurm.conf.

   invalid or expired job.