[Pw_forum] QE and mpich2, Linux

ac.rain at inbox.com ac.rain at inbox.com
Fri Dec 10 07:54:18 CET 2010


Hi Paolo,

Thank you for your assistance. that is looking better, 

This is the line I added to the script check-pw.x.j ...

PARA_PREFIX="mpiexec -n 60 -f /home/user/mpiMachinefile.txt"

when launched with "./check-pw.x.j" many 100% processes showed on each system. however I am not sure what it was doing because nothing else happened besides the text output "Checking atom...".

Usually this test completes in less than an hour on 1 core, many cores should reduce the completion time, after a couple of hours I saw nothing more than "Checking atom..." so I pressed Ctrl+C on the terminal and it printed some more information...

$ ./check-pw.x.j
Checking atom...Killed by signal 2.
[mpiexec at server4] HYDT_bscu_wait_for_completion (./tools/bootstrap/utils/bscu_wait.c:99): one of the processes terminated badly; aborting
[mpiexec at server4] HYDT_bsci_wait_for_completion (./tools/bootstrap/src/bsci_wait.c:18): bootstrap device returned error waiting for completion
[mpiexec at server4] HYD_pmci_wait_for_completion (./pm/pmiserv/pmiserv_pmci.c:352): bootstrap server returned error waiting for completion
[mpiexec at server4] main (./ui/mpich/mpiexec.c:294): process manager error waiting for completion
FAILED with error condition!
Input: atom.in, Output: atom.out, Reference: atom.ref
Aborting
$

I also tried adding "-wdir /usr/local/espresso-4.2.1/tests" at the end of PARA_PREFIX and gave the same result. I also tried uncommenting PARA_POSTFIX="./check-pw.x.j" and it gave the same result.

thanks & regards,

Nick

> -----Original Message-----
> From: giannozz at democritos.it
> Sent: Wed, 8 Dec 2010 09:42:06 +0100
> To: pw_forum at pwscf.org
> Subject: Re: [Pw_forum] QE and mpich2, Linux
> 
> 
> On Dec 8, 2010, at 8:15 , ac.rain at inbox.com wrote:
> 
>> $ mpiexec -f ~/mpiMachinefile.txt -n 10 -wdir /usr/local/
>> espresso-4.2.1/tests ./check-pw.x.j
> 
> the correct way to start the tests in parallel is to edit PARA_PREFIX
> and PARA_POSTFIX
> in the "check-pw.x.j" script
> 
> P.
> ---
> Paolo Giannozzi, Dept of Chemistry&Physics, Univ. Udine
> via delle Scienze 208, 33100 Udine, Italy
> Phone +39-0432-558216, fax +39-0432-558222
> 
> 
> 
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

____________________________________________________________
FREE 3D EARTH SCREENSAVER - Watch the Earth right on your desktop!
Check it out at http://www.inbox.com/earth



More information about the users mailing list