Update 10/21 - CEJet

Two production runs on Stampede

Run 002, model A

Jet-mass-loss-rate = 2 Msun/year, higher ambient profile (rho0 = 6.69e-9, p0 = 1e5), a0 = 49 Rsun.

Stopped at frame 28, waiting for the next queue.

Run 003, model C

Jet-mass-loss-rate = 2 Msun/year, lower ambient profile (rho0 = 1e-10, p0 = 2e3), a0 = 73.5 Rsun.

Currently running, has got to frame 5.

  • Each run is on 16 nodes (768 cores). Recent queue time on Stampede is about 3.5 days, so I submit jobs in a 3-day separation.
  • lSkipProfile = False for production quality. This takes time at every restart, (dt~13 hours between the first two chombos files). Later on each frames take about 2 - 1.5 hours, and is getting faster throughout the run time (so I didn't use the restarting at each 5 frames scheme).
  • The last run on Stampede was killed with the following message:

any ideas? 

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   PID 334217 RUNNING AT c498-002
=   EXIT CODE: 7
=   CLEANING UP REMAINING PROCESSES
=   YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
===================================================================================
   Intel(R) MPI Library troubleshooting guide:
      https://software.intel.com/node/561764
===================================================================================
TACC:  MPI job exited with code: 7
TACC:  Shutdown complete. Exiting.
Starting replacement routine
/tmp/slurmd/job4506883/slurm_script: line 69: .true.: syntax error: operand expected (error token is ".true.")

Summary of the four models planned

  • A: 2 Msun/yr jet, higher ambient, starting at 49 Rsun.
  • B: 2 Msun/yr jet, lower ambient, starting at 49 Rsun.
  • C: 2 Msun/yr jet, lower ambient, starting at 73.5 Rsun.
  • D: no jet, lower ambient, starting at 73.5 Rsun.

Comments

No comments.