MPI Continues to Run After Timeseries should terminate

[jnickla1]

All 36 of the mpi subprocesses were still running hours after the ./timeseries script successfully terminated. Log file:

************************************************************
Successfully completed generating variable time-series files
Total Time: 18680.984241962433 seconds
************************************************************
--------------------------------------------------------------------------
A system call failed during shared memory initialization that should
not have.  It is likely that your MPI job will now either abort or
experience performance degradation.

  Local host:  utley.local
  System call: unlink(2) /var/folders/rb/z2h_lx454f73ttnwhkkjdk8h0000h3/T//ompi.utley.515/pid.37919/1/vader_segment.utley.515.8ee40001.26
  Error:       No such file or directory (errno 2)
--------------------------------------------------------------------------
[utley.local:37919] 1 more process has sent help message help-opal-shmem-mmap.txt / sys call fail
[utley.local:37919] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages

Fix may be export TMPDIR=/tmp.

This only happened on one machine that I installed this code onto, the other terminated successfully and relinquished the cpu.