How to use the Star-P Software to Parallel Process on the Cranium Cluster
Star-P software allows users performing scientific, engineering or analytical computation on array or matrix-based data to use parallel architectures such as multi-core workstations, multi-processor systems, distributed memory clusters and/or utility/cloud-based environments.
The software delivers users of
MatLab and
NumPy (Numeric Python) the ability to:
- Transform serial applications for parallel deployment
- Develop and deploy new parallel applications and algorithms
Visit the official
Star-P website for more information.
To use the Star-P software: Follow the next steps. As a note, users are required to set public/private SSH authentication keys on the Cranium cluster to use the Star-P client. This allows you to SSH into any LONI machine without the need for passwords. Use this link to configure your authentication keys or send an email to support@loni.ucla.edu for assistance.
-
As the software is currently in evaluation, you need to login into one of the Cranium Cluster's compute nodes we configured to run Star-P. This machine is not accessible from all computers in the LONI network, so you need to login into a SSH proxy machine before accessing this compute node.
-
SSH into ssh.loni.ucla.edu,
-
Now, ssh into cerebro-24-114.data.cluster.loni.ucla.edu,
-
Run the command: starpClient (if you don't have /usr/local/bin in your path, then run /usr/local/bin/starpClient). You will have the following output:
-
By typing the qstat command, you can see the number of slots reserved by Star-P to run your parallel MatLab jobs in the Cranium cluster, in this case, 10 slots as shown in your Star-P client configuration file in your home directory.
--Hugo Hernandez - 16 Sep 2009