Code_Aster forums

corra · 2012-05-11 20:32:20

Hi to everybody,
I managed to set a cluster of 12 processors, and I'm trying to perform a parallel calculation. During the analysis I obtain a strange copy error:

Code:

proc.7 - seed : /mnt/shared_tmp/utente-debian01-interactif.4342/*
   COPY /mnt/shared_tmp/utente-debian01-interactif.4342/* into /tmp/utente-debian01-interactif.4342.proc.7
cp -L -r /mnt/shared_tmp/utente-debian01-interactif.4342/* /tmp/utente-debian01-interactif.4342.proc.7
                                                                        [FAILED]
Exit code : 256
cp: impossibile eseguire stat di "/mnt/shared_tmp/utente-debian01-interactif.4342/lock": gestione del file NFS interrotta

<F>_COPY_ERROR     error during copying /mnt/shared_tmp/utente-debian01-interactif.4342/* to /tmp/utente-debian01-interactif.4342.proc.3
message : cp: impossibile eseguire stat di "/mnt/shared_tmp/utente-debian01-interactif.4342/lock": gestione del file NFS interrotta

The last sentence means "NFS file management interrupted".
I obtain this error (which is not fatal, but slows the analysis down) not always for the same processor. Maybe aster-mpihosts and asrun file are useful, I attached them.
/mnt/shared is mounted in all nodes as explained in http://www.caelinux.org/wiki/index.php/ … er-Config.

Thank you in advance,
Corrado

Last edited by corra (2012-05-12 20:35:44)

Attachments:
etc.tar.gz, Size: 2,873 bytes, Downloads: 40

Thomas DE SOZA · 2012-05-13 22:17:42

Seems like a config or disk problem not related to Code_Aster. Maybe your NFS mounting is unstable or full.

TdS

corra · 2012-05-15 19:49:39

I tried to use samba network instead of nfs, but copy errors increase...maybe a syncronization problem?anyone knows how I could debug the problem?
Thank you in advance,
Corrado

Code_Aster forums

#1 2012-05-11 20:32:20

error in parallel calculation

Code:

#2 2012-05-13 22:17:42

Re: error in parallel calculation

#3 2012-05-15 19:49:39

Re: error in parallel calculation

Board footer