NDB/J

Jim Dowling, MySQL

MySQL Cluster is an in-memory, replicated, distributed database that can provide low latency response times (under 10 milliseconds) and high throughput (up to 100,000 transactions/second) on low cost hardware, customer pr. So, although you can easily configure hardware to support extremely high throughput persistent applications, there has been a perception that this efficiency often comes at the high cost of investment in developing native C/C++ applications.

Session Contents

The NDB programming concepts will be presented within the context of a new Type II Java connector (NDB/J) for cluster.

I'll show how to write and deploy a simple application in 30 minutes.

In particular, there will be a discussion of how to scale cluster applications at the application level, using NDB abstractions such as batching operations, and distribution hints to help transactions start on nodes where the data is located; as well as how to transparently handle node failure, and a general discussion on programming issues in distributed database issues.

Who should come

Anybody interested in MySQL cluster; anybody interested in high performance Java programming.

Interested in participating? Put your name down

Jim Dowling