to boot.  This should happen very rarely.

starts in accounting.

Add an srun shepard process to cancel a job and/or step of the srun process
is killed abnormally (e.g. SIGKILL).

There was an oversight in the GracTime preemption logic that resulted
in repeatedly reading the PreemptType configuration parameter and
doing a string comparison. This change will improve efficiency of
that logic.

One of our testers discovered a regression in version 2.3.1.  If a job is
pending due to PartitionNodeLimit and the limit is relieved with a
'sacctmgr modify qos name=<qos name> set flags=partitionmaxnodes' new jobs
exceeding the partition limit (but not the QOS limit) are allowed to run.
However, the pending job is never allowed to run.  Attached is a patch to
address this problem.  FYI, this problem doesn't exist in version 2.4.
Patch from Bill Brophy, Bull.

sacct --name follows the same logic as scancel/squeue, limiting the output only
to jobs having one of the specified name/s.

The job names are simply sent over the wire to be handled by the backend.
Both mysql and filetxt backends were updated to handle this new condition.

This requires an incompatible protocol change in the job_cond structure.
The perl api was also updated.

SQUEUE_NAMES was previously introduced for squeue along with
--name, but not documented.

The change is due to job's qos and account limits being re-tested more
frequently and we want to avoid logging so many messages.

Let a job be submitted even if it exceeds a QOS limit. Job will be left
in a pending state until the QOS limit or job parameters change. Patch by
Phil Eckert, LLNL.

Patch by Alexander Bersenev (Institute of Mathematics and Mechanics, Russia).

Conflicts:
	META
	NEWS