IBM says DB2 version 9 is one of the biggest upgrades to its database line in the past decade. And there is a lot of evidence to back that statement. It starts with a huge upgrade to XML support and direct XML Querying and runs to enhancing greatly accessibility and development features. In effect, IBM is saying that ESB/SOA and associated Web Services are so important that they are going to implement the ability to store and retrieve XML in its native format including important XML/Secure data encryptions and authentications. And all of this must be parsing engine tamper-proof. In effect, IBM is betting that these XML capabilities are going to be critical for database engines going forward. This review examines the new DB2 database in this light.

What's New

The Control Center's Java UI interface is the same, but notably a little faster. But most important of all it fully supports many of the new XML features including the XML Query Builder and other support wizards. In general the way that IBM has incorporated the new native XML capabilities throughout the DB2 support/utility infrastructure is one of the critical new features in this upgrade.

Db2 Viper has a number of other improvements but perhaps the next most important actually was available in DB2 version 8. That was the Express C free version of DB2. Viper continues the tradition by making available a free, full version of DB2 Viper with all the new XML, XQuery and performance improvements less the most advanced partitioning and clustering functions used largely in very big installations. This basic or Express C version of DB2 is not only available for development but also deployment free of charge. And the free deployment license allows for an unlimited number of users on dual core servers with up to 4GB of memory.

Given the advanced features of D2 Viper-this is a bargain that developers, VARs and IT organizations cannot afford to ignore. One of the most popular databases available for large scale enterprise applications can be used free of charge for development, testing and most small to medium scale deployments - easily in the 10-100 user range for many database applications. Again, this is very attractive, and not to be missed.

The new DB2 features can be classified in a number of ways targeted toward end users, feature sets, or functional attributes. We shall take the latter approach and classify the new features under four headings:
- Administration and operation enhancements
- XML Native Storage Support and integration
- New database functional features
- Runtime storage and performance improvements.
IBM claims that these new features make Viper one of the most comprehensive set of updates for DB2 in at least a decade.

Yet despite the major new XML and other improvements, the interfaces to DB2 have remained consistent. So this means that developers and DBAs will still be operating in familiar DB2 working environs. And therefore, because IBM has done such a remarkable job of integrating the new features, especially native XML, into the existing interfaces - the learning curve is reduced considerably. True with new DB2 Developers Workbench and XQuery Builder there are new features and designers to master; nonetheless IBM has made a sweeping update very approachable. So lets look at each of the DB2 updates in more detail.

Administration and Operational Enhancements

IBM is calling its improvements to automated administration Autonomic Computing. In effect the database offers, often by default, to do a number of tasks (including workflow sequences of steps) that would ordinarily be performed by a DBA or system analyst. However, the DBA can easily intervene, reconfigure and or completely takeover the tasks manually. This screenshot of the DB2 Control Center shows the direction:
Note in the lower right Data Object Info Pane some of the Autonomous tasks that DB2 manages. Some of these tasks started as early as DB2 versions 6 and 7 but many have seen enhancements through to version 9. Here are some examples of version 9's Autonomic Computing refinements:
Adaptive, self-tuning memory allocation - helps reduce or eliminate the task of configuring your DB2 server by continuously updating configuration parameters and resizing buffer pools. When enabled, this feature dynamically distributes available memory resources between different points of memory consumption. Note DBAs get to choose this option at database creation time.

Automatic storage support - automatically grows the size of your database across disk and file systems. It eliminates the need to manage storage containers while preserving the performance and flexibility of database managed storage. In DB2 Version 9.1, automatic storage support has been added for multi-partition databases. Automatic storage is enabled by default when you create new databases but DBAs can turn it off and manage it manually as required.

Automated statistics collection - called RUNSTATS, is enabled by default when you create a new database. The DB2 server collects statistical information about your data in a background process as and when required. The DB2 optimizer uses this information to ensure the most efficient retrieval of information from the database. The database stats are also used for Control Center reports. This is an example of automated admin steps.