limits

Add support for dependencies of job array elements (e.g.
"sbatch --depend=afterok:123_4 ..."). This does not support
depenendencies of ALL job array elements, only individual job
array elements.

preforms the allocation, regression in 2.6.3.

Previously the SlurmdDebug value in slurm.conf was ignored if the
previous value was not 3/init

Reorder some logis in the hostlist functions for performance improvement
specifically for "if (A & B) ..." move the fastest tests first (test
A should take less time than test B).

This avoids building hostlist information with NodeHostName and
NodeAddr information unless explisitly requested and can improve
performance for the default mode of operation by about 65%.

Correct sbatch documentation and job_submit/pbs plugin "%j" is job ID,
not "%J" (which is job_id.step_id).

Specifically setting innodb_buffer_pool_size=64
in my.conf

Without this change a job with a reason of WAIT_PART_DOWN,
WAIT_PART_INACTIVE, WAIT_PART_NODE_LIMIT, WAIT_PART_TIME_LIMIT, or
WAIT_QOS_THRES would not be cleared when that reason no longer
applied.

Add cgroup create retry logic in case one step is starting at the
same time as another step is ending and the logic to create
and delete cgroups overlaps.
bug 447

I did the merge improperly

Conflicts:
	NEWS

If a node has GRES and multiple threads per core the select/cons_res
plugin can get stuck in an infinite loop.
See bug 475
Contributed by:
PREVOST Ludovic
NEC HPC Europe

If slurmd fails to get IPMI value, then I propose to force to wait 1 second instead of asking BMC again. (Part 3/4 of the patch).
If IPMI init fails when slurmd forces to update the value, then we should not update the value. (Part 4/4 of the patch
Part 1/4 and 2/4 add a security in IPMI init because the function can be call several time.
This force to return SLURM_FAILURE if the first call failed, since the other call will not do anything.

bug 469

Previously a node failure would always requeue the job

Restore default behavior of allocating cores to jobs on a cyclic basis
across the sockets unless SelectTypeParameters=CR_CORE_DEFAULT_DIST_BLOCK
or user specifies other distribution options.
Reverts commit 7fcdc7e5
bug 466

Change Sockets to SocketsPerBoard and Procs to CPUs

If the backfill scheduler relinquishes locks and the normal job
scheduler starts a job that the backfill scheduler was actively
working, the backfill scheduler will try to re-schedule that
same job, possibly resulting in an invalid memory reference
or other badness.

Fix for --cpu_bind=map_cpu/mask_cpu/map_ldom/mask_ldom plus
--mem_bind=map_mem/mask_mem options, broken in 2.6.2.
See commit 718382da

Expect was failing periodicallly due to apparent timing problems

This messsage type:
Warning: Note very large processing time from schedule: usec=9467365 began=11:06:23.003
is reporting the end time as the began value

This prevents premature re-sending of job kill RPC
(e.g. "Resending TERMINATE_JOB request JobId=#")