karthik - 2007-09-18

Hi All,

I am trying to install pypar on a Beowulf Cluster of 3 machines. I downloaded the latest alpha version of pypar. I have mpich running on the cluster. The installation succeeded, but when i run the demo.py program, I get send/recv errors.

p1_22773: (0.007812) net_recv failed for fd = 3
p0_22762: (2.890625) net_recv failed for fd = 7
p0_22762: p4_error: net_recv read, errno = : 104
p1_22773: p4_error: net_recv read, errno = : 9
p0_22762: (4.898438) net_send: could not write to fd=6, errno = 32

I know that the communication between the nodes is good and all the configurations are right, because, when i run a simple C program with MPI_Send and MPI_Recv, it works perfectly fine, without any of these errors.

I would really appreciate it if someone could please tell me what could be going wrong here.

Regards

Karthik