[mvapich-discuss] Can you help me ?
Qi Gao
gaoq at cse.ohio-state.edu
Tue May 15 14:43:53 EDT 2007
Hi,
Thanks for trying MVAPICH2. We are glad to work with you to solve the problem.
>From the log, it seems that MPI program is having trouble to connect to MPD for CKPT support. To help us to narrow down the problem would you plesae first verify the following things?
1. Are you using the make script "make.mvapich2.ofa" to compile & install MVAPICH2 and enabling CKPT by assigning the env in the script ENABLE_CKPT to yes?
2. Are you using the MPD provided in MVAPICH2 package as your process manager? That version of MPD includes the extended functionality to support CKPT.
Thanks.
--Qi
----- Original Message -----
From: sunway qilu
To: mvapich-discuss at cse.ohio-state.edu
Sent: Monday, May 14, 2007 9:32 PM
Subject: [mvapich-discuss] Can you help me ?
Hello,
I have the same problem with the blcr and mvapich2-0.98 ,following is error which I have:
[fortest at fes01 ~]$ mpdboot -n 2
[fortest at fes01 ~]$ ps ax|grep mpd
13976 ? S 0:00 python2.3 /usr/local/mvapich2p2/bin/mpd.py --ncpus=1 -e -d
13977 pts/16 S 0:00 ssh -x -n gn01 /usr/local/mvapich2p2/bin/mpd.py -h fes01 -p 46910 --ncpus=1 -e -d
14003 pts/16 S+ 0:00 grep mpd
[fortest at fes01 ~]$ mpiexec -n 2 ./cpi
[Rank 0][cr.c: line 124]connect 24678 failed
rank 0 in job 1 fes01_46910 caused collective abort of all ranks
exit status of rank 0: killed by signal 9
following is the config file of the user fortest:
[fortest at fes01 ~]$ cat .bashrc
# .bashrc
# User specific aliases and functions
# Source global definitions
if [ -f /etc/bashrc ]; then
. /etc/bashrc
fi
export MPI_ROOT=/usr/local/mvapich2p2
export PATH=$MPI_ROOT/bin:$PATH
export MV2_CKPT_FILE=/home/fortest/ckptfile
export MV2_CKPT_INTERVAL=20
export MV2_CKPT_MAX_SAVE_CKPTS=3
export MV2_CKPT_MPD_BASE_PORT=24678
export MV2_CKPT_MPIEXEC_PORT=14678
export VIADEV_DEFAULT_TIME_OUT=16
[fortest at fes01 ~]$ cat mpd.hosts
gn01
gn02
gn03
gn04
gn05
gn06
gn07
gn08
I wait for your answers and thanks a lot.
------------------------------------------------------------------------------
_______________________________________________
mvapich-discuss mailing list
mvapich-discuss at cse.ohio-state.edu
http://mail.cse.ohio-state.edu/mailman/listinfo/mvapich-discuss
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.cse.ohio-state.edu/pipermail/mvapich-discuss/attachments/20070515/9c06543d/attachment.html
More information about the mvapich-discuss
mailing list