PrologSlurmctld, EpilogSlurmctld, MailProg, etc.

container that never had a pid added to it.

(The job ended before it began)

Previous logic never reopened the file, preventing proper functioning
of logrotate.

Conflicts:
	src/slurmctld/job_mgr.c

Use hash function to locate job records for improved performance.

Change partition write lock to a read lock as we use a different
mechanism for hidden partitions in getting individual jobs.

This is a correction in the logic of commit
3f4b2d51 on launch failures.

Make it work for more system types
Do not delete files if the build fails

here instead of cp please change this :).

Conflicts:
	src/slurmctld/job_scheduler.c

Correct logic returning remaining job dependencies in job information
reported by scontrol and squeue. Eliminates vestigial descriptors with
no job ID values (e.g. "afterany"). As depdencies are removed, the
job ID values were removed from the strings, but not the descriptors.
This eliminates both. It also checks the full job ID to make sure we do
not remove "afterany:1234" when job "123" completes.

handle any job_fail calls after the fact since it will result in deadlock
otherwise.

Conflicts:
	NEWS
	doc/man/man5/cgroup.conf.5

Add a check to make sure that the job completion RPC from a
slurmstepd match that node that the batch job is running on.
This would not be the case of for a job started on a node
if that node's slurmd fails, but the slurmstepd keeps running.
The job could then be requeued and generate a completion RPC
from both slurmstepd daemons (one per node). This logic will
ignore the job complete RPC from the node NOT currently
running the batch job.

Fix race condition on batch job termination that could result in a job exit
code of 0xfffffffe if the slurmd on node zero registers its active jobs at
the same time that slurmstepd is recording the job's exit code.
but 535