job elapsed time.

the i/o thread.

exit code.

information wasn't stored in accounting.

job/step completion.

bit_unfmt.

Signed-off-by: Danny Auble <da@schedmd.com>

HealthCheckProgram for nodes in any other states than IDLE.#978

Unload job tables rather than windows at job end. The table unload
also unloads job tables.

lost should the slurmctld restart.

Correct NumCPUs count for jobs with --exclusive option.
bug 909

This probably only happens on native Cray systems due to the
deallocation delays related to node health check. In any case,
the symptom is error message of this sort "job # dealloc of node
... bad node_offset 0 count is 0". It then fails to deallocate
the nodes GRES back for use by other jobs.
bug 973

slurm_conf_reinit.

counting as multiple nodes.

slurmstepd attempts to create it, for example left over from a previous
requeue or crash, delete it and recreate it. #961.

switch/nrt - Unload job tables (in addition to windows) in user space mode
to avoid leaking NRT.
bug 964

Fix race condition which could result in requeue if batch job exit and node
registration occur at the same time.

(From that commit)

    There was a problem when building from source where for example
    @bindir@ would resolve to ${prefix}/bin.

    This patch fixes it, based on http://www.gnu.org/software/autoconf/manual/
    autoconf-2.69/html_node/Installation-Directory-Variables.html

    It also changes opt_modulefiles_slurm to opt_modulefiles_slurm.in but
    I couldn't figure out how to get git diff to show that.

and should there be problems accessing the state files.

Fix for possible abort on change in GRES configuration.
bug 958

(commit 4cd63575)
with sacctmgr load when Parent has "'" around it

page.

special characters in their names like ':'

It seems to have been broken for some time due to logic blocking
SIGWINCH being removed from srun, which could result in the
signal being sent to any thread, not the thread designed to handle
window resizing.

failure as well as interactive user.

epilog complete message comes from that node do not process the batch step
information since the job has already been requeued because the epilog
script running isn't guaranteed in this situation.

start of requeued jobs.