OpenJPA and pureQuery revisited - technical article available
I wanted to bring your attention to an article that was recently published on developerWorks that describes the pureQuery / OpenJPA integration that I discussed in my earlier blog post, which is a new feature in WebSphere Application Server v7 that enables developers to generate SQL from their JPA application entities and named queries, which can then be bound into static DB2 packages, providing a fast path to the security and performance benefits of static access for DB2 data. In addition, when you use this capability, your JPA app can take advantage seamlessly of optimizations provided in pureQuery such as the ability to update multiple tables in a single network call.
This integration is our first step toward providing a strong, integrated stack among DB2 and WebSphere, using pureQuery as the "glue". (As an aside, the Performance Expert Extended Insight Feature also uses pureQuery to provide new insights into the interactions between Java applications and DB2, with the most capabilities being provided for Java applications in WebSphere.)
From a tooling perspective, this initial JPA/pureQuery integration is fairly light, but you will notice that the static binder utility is now invocable from the WebSphere Application Server console, so WebSphere admins won't have to switch to another tool to do the bind. In addition, you'll be able to use the new capability in Data Studio Developer 2.1 to visualizeelapsed time for SQL statements directly from the pureQuery outline.
Nevertheless, there is still much more we can do to make this process easier and less command-line driven. WebSphere plans to ship enabling technology that will "turn on" capability in Data Studio Developer 2.1 to invoke the wsdb2gen utility. In addition, you'll be able to use the output from wsdb2gen within Data Studio Developer and take advantage of other pureQuery outline capabilities, including the ability to correlate SQL statements with specific OpenJPA queries and the relations between the SQL and the associated tables and columns.
Check out the article if you get a chance.
-- Steve Brodsky