This prevented jobs from sharing nodes

remove typo

In test15.17, an salloc spawns an sbatch within that same job
allocation then exits. This results in a race condition which
can cause the terminate job RPC from slurmctld to slurmd to
briefly hang and mark the node non-responsive, causing later
tests that may need the node to fail (timeout waiting for
job allocation). A sleep is added to the salloc here to give
the batch job a chance to begin and clean up the job allocation
quickly.

This commit makes the QOS/Assocation tests a bit more thorough.
It also makes some minor format changes.

interval.

and does the same thing.

When testing times went from ~900 usec to 1.

on BGQ systems.

and SLURM_STEP_ID env vars are set.

Based upon a study of the preemption logic, putting smaller jobs
at the top of the queue (lower priority) will improve algorithm
efficiency in terms of not releasing more resources than needed.

merged together.

Conflicts:
	src/sacct/print.c

4TB of memory on them.

sacct.

The previous logic would get an authentication credential while
waiting for a message to forward. This change will re-read an
authtentication credential if the wait is too long (greater than
or equal to the message timeout).

Replace white space with tabs