Hi all
I am very much a novice at all this so please forgive my ignorance. I am
trying to set up a piece of analytical software for parallel processing of
my datasets on two quad core Mac Pros. I have managed to install OpenMPI on
the Macs, compile the software, and get it to run in parallel across the
four nodes within each Mac. My sticking point now is how to get it to run on
all eight nodes (two Macs) in parallel. The instructions on the webpage for
the software use Lam and are as follows:
*Running MrBayes*
First we have to start the 'lam' by creating a text file called lamhost. If
you are using one quad G5 machine, you need only one line "localhost cpu=4",
e.g. by typing in the terminal window
echo "localhost cpu=4" > lamhosts
When you have multiple machines, you need a line for every machine in the
lamhost file and the full machine name instead of localhost. The file might
look like
$ cat lamhosts
mac01 cpu=4
mac02 cpu=4
mac03 cpu=4
To start the lam you can type
lamboot lamhosts
from a terminal window.
To run mrbayes on *N* processors, you use
mpirun -np N ./mb
N is typically four times the number of machines you are using.
To stop the lam type
wipe lamhosts
in a terminal window.
Retrieved from "http://mrbayes.csit.fsu.edu/wiki/index.php/FAQ_3.2>
"
Obviously this doesn't work on MPI since there is no lamboot. Is there an
equivalent command to use? Or better yet, how do I go about getting these
two machines to communicate in parallel? I have scoured the web looking for
advice on this, but as I mentioned, I am new at this and I think some of the
things I have found are just way over my head right now and have hit the
point where I am giving up and asking for help. Any info/instructions/advice
(written in terms I can understand) would be greatly appreciated.
Thanks much
Ashley
|
![[Flag]](../../../../../community/mirrors/flag-us.gif){kind=small}