Conflicts:
	src/slurmctld/job_scheduler.c

Result of bad merge

Previous logic would place more tasks on each node than specified
by --ntasks-per-node, using fewer nodes than desired. This only
happens with exclusive node allocations (e.g. in partition
configuration Shared=Exclusive).

Remove stray spaces from a line

EpilogSlurmctld pthread is passed required arguments rather than a pointer
to the job record, which under some conditions could be purged and result
in an invalid memory reference.

EpilogSlurmctld pthread is passed required arguments rather than a pointer
to the job record, which under some conditions could be purged and result
in an invalid memory reference.

Conflicts:
	contribs/pmi2/pmi2_api.c

from the execution host)

Problem with previous method is if a job where the cnode failure happened
had the --no-kill option the block would be locked up until a job finished.

a block.

reboot of a block if only jobs in the list are running on it when cnodes
go into a failure state.

_p_ in the plugin function names
_g_ in the global function names

This functionality has been obsoleted by the LogTimeFormat.

Modify slurmctld message retry logic to support Cray cold-standby SDB.

Add support for "days-" (no hours)
Change local functions to have "_" prefix in their name

from "Failed to allocate resources" to
"Job submit/allocate failed"