|
[Sponsors] |
February 12, 2014, 05:22 |
Elapsed Wall Clock Time
|
#1 | |
Senior Member
Mr CFD
Join Date: Jun 2012
Location: Britain
Posts: 361
Rep Power: 14 |
Hi,
Is there a way to tell CFX to stop the simulation after 11 hours of wall clock run time? I submit jobs to a high end computer node, which the maximum allowed run time is 12 hours. Before the computer kicks me out and aborts my job I want to write a results file at the 11th hour. I thought the "Elapsed Wall Clock Time Control" option might have done this but after reading the CFX Pre guide it seems I'm wrong. I did a search, and read a comment by ghorrocks in this thread who said Quote:
Thank you |
||
February 12, 2014, 05:28 |
|
#2 |
Super Moderator
Alex
Join Date: Jun 2012
Location: Germany
Posts: 3,399
Rep Power: 46 |
I had the same problem a few years ago and as far as I remember, the elapsed wall clock time control was exactly what I needed.
|
|
February 12, 2014, 05:35 |
|
#3 |
Senior Member
Lance
Join Date: Mar 2009
Posts: 669
Rep Power: 22 |
I've had problems getting the elapsed wall clock time control to work, especially if Im using many cores for the simulation. It seems that it never kicks in, even though the wall clock time has expired. If I have to stop a simulation, I just run cfx5stop manually. Im sure there are ways to do it more efficiently, e.g. using cron as mentioned by Glenn.
|
|
February 12, 2014, 05:37 |
|
#4 |
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 17,703
Rep Power: 143 |
The elapsed wall clock time is a new(ish) feature which makes it very easy. The alternative using a cron job is the old way of doing it - but the idea of running cfx5stop to stop a job is also useful with other triggers which are not built into CFX.
As for how to use cron: http://www.lmgtfy.com/?q=cron (sorry about the sarcasm but I have been itching to use that link ) |
|
February 12, 2014, 06:02 |
|
#5 |
Senior Member
Mr CFD
Join Date: Jun 2012
Location: Britain
Posts: 361
Rep Power: 14 |
Excellent. Thanks for your replies. I'll do a bit more digging around using cron, and I'll run a few simple jobs using the command to see if it works.
P.S - Lance. I agree. I've had no luck getting the lapsed wall clock time control to work on 16 cores. I doubt it'll work when I ratchet up the power to 64 or 128 cores! |
|
February 12, 2014, 06:57 |
|
#6 |
Senior Member
Lance
Join Date: Mar 2009
Posts: 669
Rep Power: 22 |
If I remember correctly I got the control to work using 64 cores, but there was no way that it worked for 128 or more. However, cfx5stop worked fine for the same simulation. I asked Ansys about this a long time ago, but they never addressed the issue.
|
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Superlinear speedup in OpenFOAM 13 | msrinath80 | OpenFOAM Running, Solving & CFD | 18 | March 3, 2015 05:36 |
mixerVesselAMI2D's mass is not balancing | sharonyue | OpenFOAM Running, Solving & CFD | 6 | June 10, 2013 09:34 |
Water subcooled boiling | Attesz | CFX | 7 | January 5, 2013 03:32 |
same geometry,structured and unstructured mesh,different behaviour. | sharonyue | OpenFOAM Running, Solving & CFD | 13 | January 2, 2013 22:40 |
difference in CPU and wall clock time | Atit | CFX | 6 | May 22, 2007 08:43 |