Developments in ROOT I/O and trees

For the last several months the main focus of development in the ROOT I/O package has been code consolidation and performance improvements. We introduced a new pre-fetch mechanism to minimize the number of transactions between client and server, hence reducing the effect of latency on the time it takes to read a file both locally and over wide are network. We will review the implementation and how well it works in different conditions (gain of an order of magnitude for remote file access). We will also briefly describe new utilities, including a faster implementation of TTree cloning (gain of an order of magnitude), a generic mechanism for object references, and a new entry list mechanism tuned both for small and large number of selections. In addition to reducing the coupling with the core module and becoming its owns library (libRIO) (as part of the general restructuring of the ROOT libraries), the I/O package has been enhanced in the area of XML and SQL support, thread safety, schema evolution, tree queries, and many other areas. 1. ROOT I/O History ROOT I/O has been at the core of ROOT since its inception and has continually improved for 12 years. This is brief review of the major enhancements. In earlier version, the I/O was requiring the user to modify its code and was using a streamer function either automatically generated or hand coded when schema evolution was required. In 2001, we introduced support for Automatic schema evolution using the CINT dictionary information to completely drive the I/O and introduced a reference type (TRef). In 2002 and 2003 we lifted all the requirements of the user classes to be instrumented and extended all the I/O API accordingly. Since 2005 ROOT allows for files with a size larger than 2 GB. Also the compression capabilities of the I/O system for floating point umbers were improved. In 2006, we extended the TRef to work seamlessly and more efficiently with our drawing tools and improved performance over wide area networks by introducing automatic prefetching.