[MPICH] FW: a problem testing the MPICH2...no response from nodes
Rajeev Thakur
thakur at mcs.anl.gov
Mon Nov 26 11:12:57 CST 2007
-----Original Message-----
Date: Mon, 26 Nov 2007 21:02:49 +0800
From: "Horus Lee" <horus.li at gmail.com>
To: mpi-users at mcs.anl.gov
Subject: a problem testing the MPICH2...no response from nodes
MIME-Version: 1.0
Content-Type: multipart/alternative;
boundary="----=_Part_26163_24690371.1196082169278"
X-Virus-Scanned: by amavisd-new-20030616-p10 (Debian) at mailgw.mcs.anl.gov
X-Spam-Status: No, hits=0.4 tagged_above=-0.1 required=5.5 tests=BAYES_00,
HTML_10_20, HTML_MESSAGE, RCVD_BY_IP, SARE_MSGID_LONG40
X-Spam-Level:
X-MCS-Mail-Loop: mpi-users
Hello all,
Recently I'm running MPICH 1.0.6(installed by FreeBSD Ports Collection) on
bother Master (Xeon 3.0G/FreeBSD amd64/named SLIM) and Slave (Celeron
700Mhz/FreeBSD i386/named big-234), all other tests finished smoothly...but
when i use mpiexec to run example/cpi on this small cluster, the output is
[horus at slim] ~> mpiexec -l -n 2 ./cpi
0: Process 0 of 2 is on slim.xxx.xx.cn
1: Process 1 of 2 is on big-234.xxx.xx.cn
and it just stopped there without any outputs. seems like no response (the
value of PI) from both SLIM and BIG
then i use Ctrl-C to finish it. After that, I tried this command: $ mpiexec
-host big-234 example/cpi on Slim, i can get this:
Process 0 of 1 is on big-234.xxx.xx.cn
pi is approximately 3.1415926544231341, Error is 0.0000000008333410
wall clock time = 0.002024
but i can use $ mpiexec -l -n 30 /bin/hostname to get it print all the
hostnames correctly.
Can anybody help me out? All helps are greatly appreciated.
CHEERS,
HORUS Lee
More information about the mpich-discuss
mailing list