This year I added quite a few sessions to my OOW19 schedule, and I managed to go to most of them (and 1 that I wasn’t planning to go initially). As I mentioned in my OOW wrap-up post, the summary of these sessions turned out to be quite long so I separated it into a new post.
19c SQL tuning using SQLHC – this was a short session (Oracle added a 20 min theatre style sessions this year). It was nice to hear about the SQLHC tool, but I have to say that the session itself wasn’t very informative and the examples were a bit strange.
Planning a vacation using Python and ML – This presentation was simply great. Brendan Tierney talked about using machine learning with Python and cloud services to plan a trip to Scotland and decide which distilleries he should visit. He showed the technical details but used a cool story to present it.
Automatic indexing in 19c – this was presented by Richard Foote who I met personally for the first time at OOW19 (and no, he doesn’t look like David Bowie, regardless of what he says). Richard is the index master and the session was quite amazing. He managed to go through 120 slides very well (something I’ll never be able to do in 45 minutes) and most of them were examples. He said he had tried many many scenarios and put in the presentation the most interesting ones. It was very interesting and gave lots of info about how this feature works, what the risks are, when it won’t be optimal and so on. But as Richard said, this is only version 1, we expect a lot more in the future.
Active Data Guard – Pieter Van Puymbroeck (the relatively new Oracle Data Guard PM) talked about new features of Data Guard, the broker, and Active Data Guard. There are good stuff in 19c and even more in 20c, so it was a great informative session and I took some notes (maybe will have a future post about that).
Oracle RAC-ready application – this was another theatre style session, this time by Markus Michalewicz, the director of RAC PM. This was the first time I met Markus personally as well (after chatting with him quite a lot online) and he is a great guy. In this short session he talked about pitfalls that applications can hit when they are running on RAC. Markus is an excellent presenter and the session, though short, covered the topic very well.
What’s new in the optimizer – Nigel Bayliss presented new features of the optimizer in 19c and 20c. Nigel (yet another guy I met face to face for the first time at OOW19) is awesome! He is very knowledgeable and funny and the session was very interesting. The only thing I missed is more non-cloud non-exadata features. I understand that the main features are for these environments, but I really wished to hear more about the “conventional optimizer”. Maybe the “conventional optimizer” is perfect so there are no cool new things to talk about? 😉
Upgrade lab and session – I’ll talk about both together as they are on a similar topic. I first attended the hands-on lab to see how the new 19c autoupgrade works (it can upgrade you DB to 12.2 and 18c as well). This is such a cool tool, especially if you have lots of databases. The session (and the lab) was lead by Mike Dietrich (finally someone I’ve met before) and talked generally about upgrades. Both were excellent (as expected from Mike) and I got lots of valuable information about upgrades.
RAC internals – Anil Nair (man, another person I’ve met for the first time here) is a master RAC PM in Markus’s team. He is super technical and super nice and helpful and the session was the same. Lots of understanding about how things work, improvements and some new feature. The only thing that disappointed a bit was the 10-15 minutes talk from a PayPal person. He explained what they have and why they moved to RAC, etc., but I wish he would be more technical about difficulties they had, tweaks they used and other specific information about their implementation (that might help others).
Twenty features you’ll miss without Oracle – Franck Pachot (hey, I met him before OOW19) started working with PostgreSQL lately and talked about things you won’t have if you move from Oracle to other databases. He mainly talked about moving to open source (PostgreSQL and MySQL, so no SQL Server or other commercial ones) and it was interesting to see how Oracle is much more mature (the others ARE catching up, but still they have a long way to go). Franck talked about differences that will affect development (like partitions) as well as internal behavior differences (like during updates). It was really cool (as I know it would be with Franck).
Deep dive into parsing – in this session Frits Hoogland (the only one I know from Twitter and I really haven’t talked to during the conference unfortunately) showed us what parsing really is. He used debugger to add breakpoints during Oracle executions (as Frits does) and by that learned the functions, the structure and the behavior of the parse process. It was cool to see that. I was happy that I managed to follow quite easily and things made sense to me. Great way to finish this excellent conference.
Summary
I took quite a few notes from these sessions, but didn’t want to share them here as this was too much. I hope to have some time to write about some (at least) in the future.
In any case, if you have a chance to attend any of the mentioned sessions, don’t hesitate. These are excellent sessions from amazing speakers and you will learn a lot as well as have fun.