Hotsos Symposium 2008 – The Before

First of all “the before” time is over — I’m done with my presentation. It’s been the first slot of the day — 8:30 and Cary Milsap was presenting in another hall so what chances do I have to get people in? It turned out that some people actually did show up and quite a few considering the circumstances.

I have mixed feeling on the results. The presentation started very well and I managed to wake people up at the very beginning — thanks to the “equipment” I had at hands (thanks Marco and Riyaj!). You can spot one of them on the photo (thanks for the photo Marco):


The first half of the demos went pretty well but wa-a-a-y too slow — running 10g and 11g RAC on a small laptop with VMware adds some delay. I covered connection load balancing pretty well and when moved to run-time load balancing demos — things went bad. I had to skip couple demos and that screwed up the rest. I ended up showing the ideas, source code and swore it works. :) I think the audience believed me.

I needed 2 slots to cover everything I planned. But, it seems that I have the topic for the next year — it’s the second part of my presentation from this year. :) About 20 people or more came to me after presentation and expressed their full satisfaction and how much they liked the content. The fact that I overrun and didn’t finish some bits is not scary — the white paper covers it all and it’s 30+ pages filled with detailes.

Now “old” news… The best presentation of the yesterday for me was Tanel Põder’s Advanced Oracle Troubleshooting: No Magic Is Needed, Systematic Approach Will Do. He showed how to use different tools and techniques available in Oracle and OS and how they can be combined together. He showed techniques that can be used when Oracle Wait Interface is useless. The main reason I liked Tanel’s presentation is that it’s pretty much resembles the ideas of BAAG that I’m so passionate about. The main idea was using scientific methods in performance troubleshooting and it wasn’t just a simple association with a single tool, be it wait interface, Statspack or Method R, but combination of many sources of information with the idea to have a decision more informative than blind guess. Thanks Tanel – it’s great support!

There were several other great presentations but I couldn’t focus on them well enough — I had to finalize the setup of the demo and rehearse it several times (which definitely helped today but not completely). What I want to mention is Hotsos RAC Campground — the idea is an informal speaker panel in the evening. Besides me, there were James Morle, Mike Erwin and Martin Hoermann.

This was Part I and we had interesting discussion and shared our concerns and ideas. The audience was very involved and provided interesting feedback in addition to the questions. Some topics to mention — Infiniband adoption and benefits compare to Ethernet, workload partitioning, whether RAC makes your highly available for highly failable, and etc. I found confirmation to one of my observations — RAC works pretty well if you don’t push the limits of the hardware, don’t do dramatic changes to your applications and environment. But, hey do you need really RAC then? If your environment is very dynamic or hardware is loaded pretty heavily (not only CPU but IO and network as well), then stability of RAC is something very difficult to achieve. So if you need to scale out and looking for more manageability than you might want RAC but it might actually cause the opposite effect UNLESS you invest a lot. The bottom line is you either don’t need it or you pay lots and lots and lots of money to have it done properly.

Why people do RAC then? Well, Mogens Nørgaard provided 4 reason why people might decide go RAC even if they don’t need it. The second part is going to take place today and I’m looking forward to it. I have already got some questions sent from the audience and other must have collected some interesting requests as well.

That’s it for now. Sorry that it takes me too long to finish this blog post today (I’ve already got some “reminders” by email) but I’m trying to focus on the excellent presentations I have a chance to see today. Stay tuned — more to come.