Beocat Docs have now moved to http://support.beocat.cis.ksu.edu
Beocat is a medium-sized compute cluster in the CIS Department of the College of Engineering at Kansas State University. The project is overseen by Dr. Andresen and managed by Adam Tygart and Kyle Hutson.
Here we hope to be able to answer most of the questions you may have about Beocat. We strongly suggest you bookmark this page and check it regularly for updates.
Linux Basics for those not familiar with Linux
Tips and Tricks for programming and monitoring Beocat
Sun Grid Engine Information on how to submit and jobs (applications) on Beocat.
GPU Computing Information on using GPU computing.
All systems are running Gentoo Linux with a 3.10 kernel. They use PXE to boot and mount SquashFS image as the root partition. They have a local disk for swap space and some temporary storage. Home directories are accessible on all nodes via a GlusterFS volume.
We currently have 164 nodes totaling 2528 processors and 14TB+ of ram. Check the systems list for more detailed information on the hardware.
You can observe the load and status of the cluster at http://ganglia.beocat.cis.ksu.edu.
Beocat currently has the following software packages / libraries installed:
Java v1.6 and v1.7
Python 2.7, 3.2, pypy-1.9
Many different groups use Beocat for their work. For details, check out Current Projects.
Pictures have been taken of the recent reorganization: Server Room Pictures.
Information on power consumption can be found here: Power Utilization.
Some Administration scripts can be found here: Open Source administration
Nodes on Beocat aren't scheduled exclusively, by default, other people may be running jobs on the same node. Request 1 core per thread you intend to use, please DO NOT oversubscribe the nodes.
Contact and Support Information
If you have a problem, please send an email to email@example.com . This will put a service request in the queue so we can track and remedy your problem. Please keep the subject line as descriptive as possible and include any pertinent details to your problem (i.e. job ids, commands run, working directory, program versions,.. etc).