The DaytonaTM data management system is used by AT&T; to solve a wide spectrum of data management problems. For example, Daytona is managing a 94 terabyte, 7x24 production data warehouse whose largest table contains over 207 billion rows as of Nov 2003. Daytona's architecture is based on translating its high-level query language CymbalTM (which includes SQL as a subset) completely into C and then compiling that C into object code. The system resulting from this architecture is fast, powerful, easy to use and administer, reliable and open to UNIX tools. In particular, two forms of data compression plus robust horizontal partitioning and effective SPMD parallelization enable Daytona to handle terabytes with ease. Fast, large-scale in-memory operations are supported by in-memory tables and scalar and tuple-valued multi-dimensional associative arrays.

Daytona offers all the essentials of data management including a high-level query language, data dictionary, B-tree indexing, locking, transactions, logging, and recovery. Users are pleased with Daytona's speed, its powerful query language, its ability to easily manage large amounts of data in minimal space, its simplicity, its ease of administration, and its openness to other tools. In particular, Daytona supports SQL, Perl DBI, and JDBC.

Daytona, through its use in AT&T;'s call detail data warehouse, dominated the Decision Support System category of the Winter Corp Top Ten VLDB List. Daytona won First Place over all DSS entries over all platforms in the categories of Normalized Data Volume and Number Of Rows. In the Database Size category, Daytona came in second, which is not bad on the face of it, but while well-defined to be "the total disk storage used for user tables, indices and aggregates in GB", this criterion is misleading when it compares DBMS using compression with those that don't. The ones that use compression are obviously storing more information per GB of disk than those that don't, which is a good thing. In this survey, the AT&T; data warehouse is using Daytona's data compression features to great advantage whereas France Telecom is not using Oracle's compression features in any substantial way. To compare the two on this Database Size metric is to compare apples with oranges. Ideally, what one would want to do is to compare databases on the basis of the total information stored (in an information-theoretic sense). This is what Winter Corp created the Normalized Data Volume measure to do and by that measure, the AT&T; warehouse using Daytona is storing about three times more information than the second-place entry. After all, the goal is not so much to fill up a lot of disks but rather to store a lot of information.

By the way, since this AT&T; data warehouse is implemented on a Sun E10K with 400 MHz CPUs and Sun A5x00 disks, just by upgrading the hardware alone to current Sun models (like the F15K and 3510 disks), Daytona could just as easily be handling a like data warehouse three times larger, i.e., 1.5 trillion records and 300 terabytes of uncompressed data and indices.

  1. Compared To Awk & Perl For Flat File Processing
  2. Instructional Queries
  3. Recently Added Features

For further information on Daytona, please send email to .

AT&T; is the sole source for the Daytona product, service and support and is the only company authorized to use the Daytona trademark for a database product.

home | people | projects | research areas | resources | search

Copyright and Privacy Notice