The problem is the WiFi connection!

We were connecting the computers using WiFi, we have changed to a cable
connection and the program is working.

It seems that the port that MPI_Send and MPI_Recv use is closed.

Do you know which port use these commands?

Thank you,

Sofia
> On Sep 30, 2008, at 12:02 PM, Sofia Aparicio Secanellas wrote:
>
>>> 1. From prior mails, you do not seem to have iptables running to
>>> block any ports -- is there any other port blocking software
>>> running, perchance?
>>
>> I do not think so. What can I do to be sure that I do not have any
>> other port blocking software running?
>
> Test that random TCP ports can connect to each other.  Perhaps
> something like this on computer 1:
>
> computer1$ nc -lp 123456
>
> Then on your other computer:
>
> computer2$ telnet computer1 123456
>
> If the "telnet" connects, then you should be able to type anything and
> have it appear on computer1 (where nc) is running.  Hit Ctrl-C in the
> window where "nc" is running to kill the connection.
>
> If this works, then you likely do not have port blocking software
> running.
>
>> How can I do to attach the MPI processes with a debugger?
>> I have used before:
>> dbx - PID
>> where
>> It seems that the results that I have obtained were not very helpful.
>
>
> Ah yes, sorry, this thread has been so long that I missed that.  I see
> the stack traces, and like Terry said, it shows that the sends and
> receives are pending, but not *why*.  Well, let's see the results of
> the "nc" test from above before deciding how to move onward...
>
> --
> Jeff Squyres
> Cisco Systems
>
> _______________________________________________
> users mailing list
> us...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/users
>


Reply via email to