Clustered File System

Once you start to do some serious computing on your Virtual Supercomputer you will realize that the filesystem is a bit of a bottleneck for many of your applicatons. Large computer clusters place an unbelieveable load on storage systems. Most storage solutions just weren't built to handle thousands of concurrent reads and writes and the bigger your cluster becomes the more demanding your storage requirements become.

But there's a simple and inexpensive solution to this dilemma... Lustre. Lustre is a massively parallel distributed file system, generally used for large scale cluster computing. As are most of the tools we recommend, Lustre is open source GPL'ed software that you can run on comodity hardware. Many of the systems in the Top500 list use Lustre based storage systems.

In our second publication "The ClusterChimps Guide to Clustered File Systems" you will find all the information you need to build a clustered file system for your Virtual Supercomputer. This book is divided into two sections. In the first section we discuss different types of network attached storage providing you with high level overviews of the most popular. In the second section we focus extensively on Lustre showing you how you can build an inexpensive high performance NAS system for your Virtual Supercomputer.

The ClusterChimps Guide to Clustered File Systems
Due out in the fall of 2012... assuming the world doesn't end.



Code Examples
This tarball contains the code for the examples in the book above.

ClusterChimps © 2009