[mpich-discuss] A runtime environment variable tocontroleager/rendezvous protocol?

Rajeev Thakur thakur at mcs.anl.gov
Mon Mar 8 23:54:29 CST 2010


If you are using the default nemesis channel in MPICH 1.2.1, I believe
the eager/rendezvous threshold is set by the value
MPIDI_CH3_EAGER_MAX_MSG_SIZE in
src/mpid/ch3/channels/nemesis/include/mpidi_ch3_post.h. You can change
it there and then run make clean, make, make install in the top-level
directory. Not sure if it can be changed via an env variable.
 
Rajeev


  _____  

From: mpich-discuss-bounces at mcs.anl.gov
[mailto:mpich-discuss-bounces at mcs.anl.gov] On Behalf Of Bin Bao
Sent: Monday, March 08, 2010 7:05 PM
To: mpich-discuss at mcs.anl.gov
Subject: Re: [mpich-discuss] A runtime environment variable
tocontroleager/rendezvous protocol?


I am actually using MPICH2 1.2.1, and don't have MPICH installed yet. I
was looking for the environment variable that gives me the control to
the eager/rendezvous protocol, but only found "MPICH_LONGVLONGTHRESH" in
the MPICH document, "Installation and User's Guide to MPICH, version
1.2.7".

Bin


On Mon, Mar 8, 2010 at 6:03 PM, Rajeev Thakur <thakur at mcs.anl.gov>
wrote:


Which version of MPICH are you using and what device?
 
Rajeev


  _____  

From: mpich-discuss-bounces at mcs.anl.gov
[mailto:mpich-discuss-bounces at mcs.anl.gov] On Behalf Of Bin Bao
Sent: Monday, March 08, 2010 2:41 PM
To: mpich-discuss at mcs.anl.gov
Subject: [mpich-discuss] A runtime environment variable to
controleager/rendezvous protocol?


Hi,

Is there a runtime environment variable in MPICH2 that can be used to
control when to switch from eager protocol to rendezvous protocol? It
seems that "MPICH LONGVLONGTHRESH" in MPICH can do this.

Bin



_______________________________________________
mpich-discuss mailing list
mpich-discuss at mcs.anl.gov
https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/mpich-discuss/attachments/20100308/98cfe860/attachment.htm>


More information about the mpich-discuss mailing list