When we run some MPI programs started using mpiexec, MPI system complains:
mpiexec_XiAn-172-26-3-109 (mpiexec 392): no msg recvd from mpd when expecting ack of request
My colleague Rick Jin found the following explanation using Google:
This error may deal with the follwing reasons:Our environment has no high performance network. I larger the value of MPIEXEC_RECV_TIMEOUT in the mpiexec.py file and solve this issue.
In mpiexec.py, variable MPIEXEC_RECV_TIMEOUT is set to 20 by default. We can change the value using command line parameter -recvtimeout. For example:
mpiexec -recvtimeout 100
Many thanks to Rick!