MPI RUN

Topics about the HYSPLIT trajectory model.
Post Reply
Smaug
Posts: 55
Joined: July 1st, 2017, 8:28 am
Registered HYSPLIT User: No

MPI RUN

Post by Smaug » August 17th, 2017, 7:16 am

Hi,

I am currently having difficulty using MPI for a trajectory run. I've changed the executable call to:

mpirun -np 4 {HOME}/...../hytm_std

This runs the model but with only one core and then returns a segmentation fault at the end. MPI runs have been successful with concentration runs.


Thanks

ariel.stein
Posts: 642
Joined: November 7th, 2012, 3:14 pm
Registered HYSPLIT User: Yes
Contact:

Re: MPI RUN

Post by ariel.stein » August 17th, 2017, 5:35 pm

How many trajectories are you calculating?

Smaug
Posts: 55
Joined: July 1st, 2017, 8:28 am
Registered HYSPLIT User: No

Re: MPI RUN

Post by Smaug » August 18th, 2017, 4:35 am

I am calculating 96 trajectories. Three trajectories for each sampling entry in the sampling input file.

ariel.stein
Posts: 642
Joined: November 7th, 2012, 3:14 pm
Registered HYSPLIT User: Yes
Contact:

Re: MPI RUN

Post by ariel.stein » August 18th, 2017, 3:48 pm

Can you please send me your CONTROL file?

Post Reply