like the switch plugin functions. When a batch job runs a core
specialization job on the same node, it was failing because both
tried to use specialized cores at the same time.

Conflicts:
	slurm/slurm_errno.h
	src/common/slurm_errno.c

A configuration change trigger event occurs when a node state
changes (e.g. Up, Down, Drain, etc.)

This reverts commit f3a66dd8.

This reverts commit 421185f4.

plugin, which call corresponding alpscomm functions. This enables job
steps to give up some of their network resources on suspend, increasing
the amount of resources which can be given to each job.

Attempt to create duplicate event trigger now generates ESLURM_TRIGGER_DUP
("Duplicate event trigger").

Modify strigger to accept arguments to the program to execute when an
event trigger occurs.

Added strigger option of -N, --noheader to not print the header when
displaying a list of triggers.

batch jobs have cpus_per_task set to zero, which resulted in an
error of "task/cgroup: task[0] unable to set taskset '0x0'"

Recover a list of running jobs when slurmd restarts. This job
list is used to determine when a job suspend can take place.
This patch also adds job suspend suspend/retry logic since
a job suspended immediately after launch can (briefly) return
an error that is it not ready yet.

If the job to be suspended can not be found, then proceed with the
suspend RPC after 3 seconds, Previous logic would hang indefinitely.
The job would not be found if it was launched, then the slurmd restarted.
The record keeping for launched jobs after slurmd restarts is fixed
in a separate patch.

This reverts commit bf83bcf7.

This reverts commit 4294c561.

requeue, requeuehold and release operations.

added is a user process or not.

cont_id contained in it.

need to do.

Do not purge the script and environment files for completed jobs on
slurmctld reconfiguration or restart (they might be later requeued).
Purge the files only when the job record is purged.
bug 834

If a job --mem-per-cpu limit exceeds the partition or system limit, then
scale the job's memory limit and CPUs per task to satisfy the limit.
bug 848

not finished yet otherwise if requeued the job may enter an invalid
COMPLETING state.

messages, moving some messages to higher debug levels, and
consolidating some needlessly verbose messages.

applications from the minimum (1%) to the more appropriate 25%.

the Aries network.

network systems

- Resolves a bug involving nested batch steps and aeld.

(mostly useful for debugging, it isn't used for much)

Sets task_is_app, which tells the kernel that this is an application
process (for core specialization)

limits to 4 concurrent jobs per node for our network resources.