I was connecting the two systems with out a switch. Would this have anything to do with it,
i loook at the system error messages and i get the following
ib0: no IPv6 routers present
ib1: no IPv6 routers present
ib1: no IPv6 routers present
My eth subnet is 10.1.1.1
/sbin/ifconfig ib0 10.255.255.4 netmask
255.0.0.0 and so forth for both the systems.
Now i can ping from one system to the other using the ib addresses. How ever when i run the bandwidth script i get the error. I was under the assumption that this should work. Is this setup correct.
Sicerly
Rangam
On 7/14/05, Tziporet Koren <[EMAIL PROTECTED]> wrote:
Its means that you have some connectivity problems thus got retry exceeded.Check your cables or see if the switch dropped packets.Tziporet-----Original Message-----
From: Srirangam Addepalli [mailto: [EMAIL PROTECTED]]
Sent: Wednesday, July 13, 2005 9:52 PM
To: openib-general@openib.org
Subject: [openib-general] VAPI_RETRY_EXC_ERR, vendor code=81Hello All,I was trying to test bandwidth using the following command/usr/local/topspin/mpi/mpich/bin/mpirun_ssh -np 2 minigar c0-0 /usr/local/topspin/mpi/mpich/bin/mpi_bandwidth 1000 262144How ever i get the following error messages./usr/local/topspin/mpi/mpich/bin/mpirun_ssh -np 2 c0-1 c0-0 /usr/local/topspin/mpi/mpich/bin/mpi_bandwidth 1000 262144
Got completion with error, code=VAPI_RETRY_EXC_ERR, vendor code=81
mpi_bandwidth: mpid/ch_vapi/viacheck.c:2109: viutil_spinandwaitcq: Assertion `sc->status == VAPI_SUCCESS' failed.
bash: line 1: 6813 Aborted /usr/bin/env MPIRUN_HOST=c0-1 MPIRUN_PORT=34474 MPIRUN_RANK=1 MPIRUN_NPROCS=2 MPIRUN_ID=13416 MPIRUN_NODENAME=c0-0 /usr/local/topspin/mpi/mpich/bin/mpi_bandwidth 1000 262144
Timeout alarm signaled
Cleaning up all processes ...done.
Killed by signal 15.
why do i get this error.Rangam
_______________________________________________ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general