to avoid deallocation when job failed before actually starting.

Not fully implemented.

previously excluded all parttions that overlapped with one that had pending jobs that needed
    resources.
now exclude only partitions that are a superset of the partition with pending jobs

has been satisfied.

    to zero (possibly changing a pending job's resource requirements from a 
    processor on some node to the full node).

the node list, count of allocated nodes, and number of processors allocated per node