[mpich-discuss] Large Message Sizes with Mpich2 V 1.0.8

[Arsenault, Benoit]

UNRESTRICTED | ILLIMITÉ 

Rajeev,
 
Similar problems were experienced by other MCNP users.  The feedback that I received from a user was to re-compile the code with a modified version of mpich2 in order to handle large messages.  I browsed the documentation for mpich2 and haven't found any limitation that would explain why my runs crash when I use large models.
 
Since yesterday I modified the MCNP source code for debugging purposes and realized that the crash happens when the send buffer is being initialized.  I assume that MCNP is trying to send out large amount of information and requires significant resources on the system.  The amount of memory consumed by the system is larger than the RAM + swap space.
 
I will try to increase the swap space on my system and hopefully this will help me out to narrow down the problem.
 
Thanks
 
Benoit

-----Original Message-----
From: mpich-discuss-bounces at mcs.anl.gov [mailto:mpich-discuss-bounces at mcs.anl.gov]On Behalf Of Rajeev Thakur
Sent: March 16, 2009 5:03 PM
To: mpich-discuss at mcs.anl.gov
Subject: Re: [mpich-discuss] Large Message Sizes with Mpich2 V 1.0.8


What maximum message predefined message size in MPICH2 are you refering to?
 
Rajeev


  _____  

From: mpich-discuss-bounces at mcs.anl.gov [mailto:mpich-discuss-bounces at mcs.anl.gov] On Behalf Of Arsenault, Benoit
Sent: Monday, March 16, 2009 3:21 PM
To: mpich-discuss at mcs.anl.gov
Subject: [mpich-discuss] Large Message Sizes with Mpich2 V 1.0.8



UNRESTRICTED | ILLIMITÉ 

Mpich2 Users, 

I compiled the 64-bit mpich2 Version 1.0.8 with MCNP5 Version 1.40 on Linux RedHat.  I can run reasonable size problems without any difficulties, but the code crashes when I use large models.  The error message that I get is listed below and relates to a subroutine in MCNP5.

 [pe:0] **DOTCOMM Error** DOTCOMMI_PACK (dotcommi_pack.c:104[!(( dotcommp_sbuf.data != ((void *)0) ))]) (alloc(dotcommp_sbuf.data)) (2069154944)

 [pe:0] **DOTCOMM Error** DOTCOMMI_PACK (dotcommi_pack.c:104[!(( dotcommp_sbuf.data != ((void *)0) ))]) (alloc(dotcommp_sbuf.data)) (2049205848)

 [pe:0] **DOTCOMM Error** DOTCOMMI_PACK (dotcommi_pack.c:104[!(( dotcommp_sbuf.data != ((void *)0) ))]) (alloc(dotcommp_sbuf.data)) (2029256752)

 [pe:0] **DOTCOMM Error** DOTCOMMI_PACK (dotcommi_pack.c:104[!(( dotcommp_sbuf.data != ((void *)0) ))]) (alloc(dotcommp_sbuf.data)) (2029256704)

 [pe:0] **DOTCOMM Error** DOTCOMMI_PACK (dotcommi_pack.c:104[!(( dotcommp_sbuf.data != ((void *)0) ))]) (alloc(dotcommp_sbuf.data)) (2028987040)

 [pe:0] **DOTCOMM Error** DOTCOMMI_PACK (dotcommi_pack.c:104[!(( dotcommp_sbuf.data != ((void *)0) ))]) (alloc(dotcommp_sbuf.data)) (2028987032)

 [pe:0] **DOTCOMM Error** DOTCOMMI_PACK (dotcommi_pack.c:104[!(( dotcommp_sbuf.data != ((void *)0) ))]) (alloc(dotcommp_sbuf.data)) (2028592456).

I browsed the MCNP Users' group and similar problems have been encountered where users had to re-compile mpich2 with a change to increase the maximum size of a message.  I couldn't find a way to go beyond the maximum size of a message pre-defined in mpich2 V 1.0.8.

Would anyone knows if mpich2 V 1.0.8 has the same limitations in terms of maximum message size, and if this can be resolved?


Thanks for your help. 


Ben 





CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE

This e-mail, and any attachments, may contain information that
is confidential, subject to copyright, or exempt from disclosure.
Any unauthorized review, disclosure, retransmission, 
dissemination or other use of or reliance on this information 
may be unlawful and is strictly prohibited.  

AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE

Le présent courriel, et toute pièce jointe, peut contenir de 
l'information qui est confidentielle, régie par les droits 
d'auteur, ou interdite de divulgation. Tout examen, 
divulgation, retransmission, diffusion ou autres utilisations 
non autorisées de l'information ou dépendance non autorisée 
envers celle-ci peut être illégale et est strictement interdite.	

CONFIDENTIAL AND PRIVILEGED INFORMATION NOTICE

This e-mail, and any attachments, may contain information that
is confidential, subject to copyright, or exempt from disclosure.
Any unauthorized review, disclosure, retransmission, 
dissemination or other use of or reliance on this information 
may be unlawful and is strictly prohibited.  

AVIS D'INFORMATION CONFIDENTIELLE ET PRIVILÉGIÉE

Le présent courriel, et toute pièce jointe, peut contenir de 
l'information qui est confidentielle, régie par les droits 
d'auteur, ou interdite de divulgation. Tout examen, 
divulgation, retransmission, diffusion ou autres utilisations 
non autorisées de l'information ou dépendance non autorisée 
envers celle-ci peut être illégale et est strictement interdite.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/mpich-discuss/attachments/20090317/1e846dd8/attachment.htm>