A NEW SCALABLE MULTI-NODE EVENT LOGGING SYSTEM FOR BABAR

The BaBar experiment is currently operating near the rate limit of its ability to log event data to disk and tape using the existing hardware and software systems. Consequently we have chosen to design and implement a new system for logging event data. The new system is designed to be scalable, so that the data rate can be increased by adding systems at one of three levels. It also has the property that data can be logged at almost unlimited burst rates without introducing dead time. The key to these features lies in the use of many nodes within the level three trigger system of BaBar. This allows the events to first be logged to local disks within the trigger system, and then later to be merged to any of multiple merge servers in non-real-time.