The other day I mentioned my goals for attendees for my upcoming ODTUG presentation on Drillbridge. Today I’m going to talk about my goals for my presentation on Oracle Data Integrator (ODI).

Over the last few years I have presented on ODI a handful of times. My main presentation on it has been highlighting a success story with it, where ODI was used to clean up and facilitate a lot of automation and ETL jobs for a health services company that [as you can imagine] has tons of data flying around everywhere. This previous presentation was more of a high-level affair, where I talked very generically about what the benefits of ODI were over the previous solution. Wanting to add a little more technical meat to the presentation, I appended what started off as a small section at the end where I take a look at just how ODI works under the covers.

While the “business” or high-level part of the presentation was all well and good, I found myself getting really excited to explain just how awesome the workings of the details of ODI ETL jobs were, and what started out as a 10-minute flight of fancy into the lower depths of something technical has now been promoted, as it were, to full on presentation.

In other words, I am going to spend an entire presentation literally tearing apart a single ODI interface, explaining why the steps are sequences the way they are, how to build idiomatic ODI interfaces, the affect of various options (journalization, query hints, delete all vs. truncate, etc.), update strategies, and more. I’m also going to marry up what is ostensibly an IT tool with various concepts from computer science, including the notion of idempotence (and why it’s a good thing).

With any luck, the attendee coming to this presentation will have a new or expanded understanding of how interfaces/mappings work, feel comfortable with modifying knowledge modules, and learn a trick or two when it comes to debugging ODI jobs. This will be a nuts and bolts, technical deep dive. While it’s ostensibly an advanced content presentation, I believe that people with even only a cursory familiarity should benefit from it as well. If you haven’t worked with ODI at all but are curious (and intrepid!) about what it can do, I think you’ll also benefit. So to all of my pure-Hyperion colleagues that haven’t dipped their toes in the ODI pool just yet, this is a great chance to hop on a different track and expand your horizons – I hope to see you there!