parallel switching

Mohsin · August 19, 2010, 21:31

Is it ok to switch between Parallel processing and non parallel processing. As with parallel processing the SWAP and moothing feature is not applicable so for swaping and smoothing feature can we work withnon parallel processing and for iterations switch to parallel processing.

would it have any effect on calculations and solution?

Thanks for your help.
Mohsin

-mAx- · August 20, 2010, 01:41

you can run a job in parallel, after x iterations you can write case and date.
Then ready the case and date in serial and re-iterate.
Take care always writing case

Mohsin · August 20, 2010, 01:44

Max thanks for ur reply

After 5000 iterations (when i got convergence) I changed to non parallel solver and start iterations again. there was a drastic rise in residuals but they reduced to convergence after some iterations.

I hope i m doing the same which u said here.

-mAx- · August 20, 2010, 03:26

yes I experienced it too. (just take care writing data file AND case file together.
Because if you read case (saved from parallel) and data file (saved from serial), then you get weird results.
Regarding the residuals drop it is ok. You get the same if you change any numerical schema or whatever

August 19, 2010, 21:31	parallel switching	#1
Mohsin Senior Member Mohsin Mukhtar Join Date: Mar 2010 Location: South Korea Posts: 249 Rep Power: 17	Is it ok to switch between Parallel processing and non parallel processing. As with parallel processing the SWAP and moothing feature is not applicable so for swaping and smoothing feature can we work withnon parallel processing and for iterations switch to parallel processing. would it have any effect on calculations and solution? Thanks for your help. Mohsin

August 20, 2010, 01:41		#2
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	you can run a job in parallel, after x iterations you can write case and date. Then ready the case and date in serial and re-iterate. Take care always writing case __________________ In memory of my friend Hervé: CFD engineer & freerider

August 20, 2010, 03:26		#4
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	yes I experienced it too. (just take care writing data file AND case file together. Because if you read case (saved from parallel) and data file (saved from serial), then you get weird results. Regarding the residuals drop it is ok. You get the same if you change any numerical schema or whatever __________________ In memory of my friend Hervé: CFD engineer & freerider

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Script to Run Parallel Jobs in Rocks Cluster	asaha	OpenFOAM Running, Solving & CFD	12	July 4, 2012 22:51
Performance of GGI case in parallel	hannes	OpenFOAM Running, Solving & CFD	26	August 3, 2011 03:07
HP MPI warning...Distributed parallel processing	Peter	CFX	10	May 14, 2011 06:17
IcoFoam parallel woes	msrinath80	OpenFOAM Running, Solving & CFD	9	July 22, 2007 02:58
Parallel Computing Classes at San Diego Supercomputer Center Jan. 20-22	Amitava Majumdar	Main CFD Forum	0	January 5, 1999 12:00

August 20, 2010, 01:44		#3
Mohsin Senior Member Mohsin Mukhtar Join Date: Mar 2010 Location: South Korea Posts: 249 Rep Power: 17	Max thanks for ur reply After 5000 iterations (when i got convergence) I changed to non parallel solver and start iterations again. there was a drastic rise in residuals but they reduced to convergence after some iterations. I hope i m doing the same which u said here.