describing that jobs must be drained from cluster before deploying
any checkpoint plugin.

Re-order existing code so that per-cpu memory limits will be
enforce with both cpus_per_task and ntasks_per_node limits.

Properly enforce a job's cpus-per-task option when a job's allocation is
constrained on some nodes by the mem-per-cpu option.
bug 590

This just shuts down the underlying select plugin used by select/cray
in order for it to free all of its allocated memory

bug 586

Set GPU_DEVICE_ORDINAL environment variable.

Previous logic would try to pick a specific node count and on a
heterogeneous system, this would cause a problem. This change
largely reverts commit a270417b

starts and then a minor typo fix

i.e. salloc -n32 doesn't request the number of nodes and with the previous
code if this request used 4 nodes and only 1 was left in GrpNodes it
would just run with no issue since we were checking things before we
selected how many nodes it ran on.

Now we check this afterwards so we always check the limits on how many
nodes, cpus and how much memory is to be used.

Fix step allocation when some CPUs are not available due to memory limits.
This happens when one step is active and using memory that blocks the
scheduling of another step on a portion of the CPUs needed. The new step
is now delayed rather than aborting with "Requested node configuration is
not available".
bug 577

For Kelly ;)

No change in logic, just rename a variable for better clarity.

based on ionode count correctly on slurmctld restart.

would be set, but this is safer.