Hello All,
 
I am a very new to cluster work and stuff and I think the question that i am about to ask is very simple.
I have made up a cluster using MPICH2.I have like 9 nodes. but I am first testing on jst 2 of them
I have installed MPI on both of the nodes and ran an mpd on one of them.
 
############################################

[EMAIL PROTECTED] ~]$ mpd &

[1] 13420

The number 13420 is given back as the PID for MPD, it is the signal that MPD was launched successfully.

Now type mpdtrace –l (it is the non capital English letter "L" not the number "1") at the command prompt.

[EMAIL PROTECTED] ~]$ mpdtrace -l

n01.bw01.giki.edu.pk_34080 (192.168.0.1)

The number 34080 is the port number. Keep it for the future reference while launching mpd daemons on the slave nodes. Now type ssh n02 on the command prompt.

[EMAIL PROTECTED] ~]$ ssh n02

After being logged in, type

[EMAIL PROTECTED] ~]$ mpd -h n01 -p 34080 -n &

[EMAIL PROTECTED] ~]$ mpdtrace –l

This would print the following response

n01.bw01.giki.edu.pk_34080 (192.168.0.1)

n02.bw01.giki.edu.pk_33295 (192.168.0.2)

###################################

all of this seems to be fine i guess.

Now i run a simple example program using

mpirun  ./cpi

and i get this..

Process 0 of 1 is on n01.bw01.giki.edu.pk

 

BUT when i do this

mpirun -np 2 ./cpi

The thing is kind of stuck...there is no proceedings.....

Now i dont know where the problem might be and what mst i do.Any help in this regard would be highly appreciable.

 

Regards

Fahad



Express yourself instantly with MSN Messenger! MSN Messenger Download today it's FREE!
_______________________________________________
Beowulf mailing list, [email protected]
To change your subscription (digest mode or unsubscribe) visit 
http://www.beowulf.org/mailman/listinfo/beowulf

Reply via email to