Add logic to sleep and retry if slurm.conf can't be read.
Without this, the slurmd daemons may die and when the SlurmdTimeout
is reached, the nodes will be marked DOWN and their jobs will be
killed.
In the long term, it would be good to exit only if the read files
on program startup, and the daemons keep running with old configuration
on reconfiguration, but I don't have time to do that work now.

Fix invalid memory reference if script returns error message
for user. Previous code failed to set static variable to NULL
resulting in xfree of memory previously freed elsewhere.

Add support for job array options in the qsub command, in #PBS
options for sbatch scripts and set the appropriate environment
variables in the spank_pbs plugin (PBS_ARRAY_ID and PBS_ARRAY_INDEX).
Note that Torque uses the "-t" option and PBS Pro uses the "-J"
option.

slurm.conf.  Rebooting daemons after adding nodes to the slurm.conf
is highly recommended.

Add a job flag to indicate when the EpilogSlurmctld us running
and don't purge the job record until it completes. This lets
the EpilogSlurmctld requeue the job and otherwise manage it.
bugs 635 and 636

bug 640

application environment the PMI2 library will connect the the PMI2
server (slurmstepd) instead of using the provided PMI_FD socket.

Cache results for major performance improvement.
bug 636

If a job request explicitly requests a GRES count of zero and that
is not the last GRES in the slurm internal data structures, the job
request will be rejected.
bug 633

Perhaps should also look into doing this for nodeinfo and libstate

this would cause pmd's to hang.

code.

interval.

and SLURM_STEP_ID env vars are set.

merged together.

4TB of memory on them.

sacct.

Add SchedulerParameters options of preempt_reorder_count and
preempt_strict_order.

This is for consistency. All of the other backfill parameters
start with "bf_".