Instead of going to the normal OpenWorld events today I went to OakTable World. Now my brain is fried from information overload. 🙂
It started at 8 am with a nice talk about hash joins and Bloom filters. Toon Koppelaars had some nice moving graphics showing how bloom filters work. I’ve studied Bloom filters before but I’m not sure I understood it with the clarity that I had after this talk.
Then I did my talk at 9 am. The best part for me was that we had a number of questions. I ended up skipping several slides because of time but I felt like we helped people get what they wanted out of it by having the questions and discussion. In retrospect my talk could have used more of an introduction to Delphix itself for this audience but I think we covered the essentials in the end.
Next Kellyn Pot’Vin-Gorman did more of a community service type of talk which was a change of pace. She had a Raspberry Pi project which was a stuffed bear that would take your picture and post it on Twitter. It was an example of the type of project that kids could do to get them interested in computer technology.
My brain began to turn to mush with Marco Gralike’s talk on XML and 12c In-Memory Column store. I’m sure I’m absorbed something but I’m not that familiar with Oracle’s XML features. Still, at least I know that there are in memory features for XML which I can file away for the future.
Several amusing 10 minute Ted talks followed. Most notable to me was Jonathan Lewis’ talk about how virtual columns and constraints on virtual columns could improve cardinality estimates and thus query performance.
Cary Millsap talked about a variety of things including things like what he covered in his book. He shared how he and Jeff Holt were hacking into what I assume is the C standard library to diagnose database performance issues, which was pretty techy.
Gwen Shapira’s talk on Kafka was a departure from the Oracle database topics but it was interesting to hear about this sort of queuing or logging service. Reminds me in some ways of GGS and Tibco that we use at work but I’m sure it has different features.
Alex Gorbachev gave a high level overview of Internet of Things architectures. This boiled down to how to connect many possibly low power devices to something that can gather the information and use it in many ways.
Lastly, we went back to the Oracle database and my brain slowly ground to a halt listening to Chris Antognini’s talk on Adaptive Dynamic Sampling. I had studied this for my OCP but it has slowly leaked out of my brain and by 4 pm I wasn’t 100% efficient. But, I got a few ideas about things that I can adjust when tuning this feature.
Anyway, brief overview. I’m back to normal OpenWorld tomorrow but it was all OakTable today. It was a good experience and I appreciated the chance to speak as well as to listen.