OUGN Spring Seminar 2013
Posted by Martin Bach on April 21, 2013
The annual conference held by the Oracle User Group in Norway has once again been just great. It was the second year I went and I have to admit that it was every bit as good as last year, and that’s holding a very high standard.
The combination of such great hosts, great speakers and a wonderful atmosphere make this one of the best conferences to attend in Europe. The added benefit of being on a boat makes it a great opportunity to meet the speakers and to hang out with during dinner and after the sessions. Unfortunately I had to leave a day early and write these lines while on a train back home.
So how has it been? Unlike last year I flew in on the day of the event on Wednesday morning into Oslo together with Martin Widlake. Martin Nash and I successfully pitched RAC Attack II to the event organizers and Martin started off in his usual stride (I am referring to Martin Nash here when mentioning “Martin” by the way). There were a few attendees less this year compared to the first RAC Attack we did but Martin has prepared cool stuff to show those who came to the sessions.
We had support from Frits Hoogland and Bjorn Rost at various parts of the event which lead to a lot of interesting discussions around workload management in RAC. While we were overcoming a network particularity with the hotel’s wireless network I used a parts of my “Grid Infrastructure and High Availability Deep Dive” seminar to demonstrate the message we wanted to put across: workload management is _the_ way to consolidate applications on RAC. It is also the only way to allow for rolling patches of infrastructure and applications. What good is RAC if you are experiencing downtime in your application because of restarts of the application servers? Sadly I have not come across an application yet that was truly RAC-aware. Often though you see FUD take over and the benefits of RAC cannot be used.
Martin went into a lot of detail in his preparations to demonstrate the usefulness of using Oracle’s Universal Connection Pool (UCP) as well as .Net for Fast Connection Failover (FCF). In my very personal opinion anyone seriously considering RAC for availability definitely has to spend time on FCF, or otherwise save the license and use an active/passive cluster or RAC One Node instead. With the increasing importance of Exadata and the Oracle Database Appliance RAC awareness is really needed, especially with developers. In my opinion it is not enough to simply use Transparent Application Failover and stop there-TAF is so 1990s! When using TAF then only as a stopgap measure until the application has been properly ported to RAC.
I would like to thank the attendees for their support and the great time I had. It was particularly cool to put faces to twitter handles, like in the case of Philippe Fierens who sat in the lab.
There are a few things to take away and discuss with the rest of the RAC Attack organisers such as the move away from VMWare Server which doesn’t really work with Windows > XP. Other than that I still believe in the format and want to thank Jeremy Schneider who initiated it.
Thursday we boarded the boat to start the cruise to Kiel which gives you plenty of time to chat with people. The conference started great with many interesting sessions to attend. I would like to give credit to Joel Goodman, who presented in quite rough waters as if it were plain sailing. His subject was Resource Management and RAC and I have not yet had such a precise explanation of global enqueue and buffer cache handling in RAC before. Should he repeat this talk at another user group you should definitely go as it is so worth it.
There were other great sessions Martin Nash already covered in his summary so if you are interested in knowing what you missed, go over to his blog post now.
Friday morning it was time for my talk about emerging trends in storage and from my personal point of view I was happy with it. Of course the feedback forms will tell me if I’m right with my gut feeling. I then went to see Doug Burns with his talk about 10053 trace files and really liked it. But then I like the way Doug presents anyway, and he is one of these speakers you want to watch out for on any conference agenda. After this I had to leave the boat since we had arrived in Kiel in the meantime.
The only regret I have was not being able to make it for the dinner on Thursday evening but my body felt in no condition to consume food or drinks. I wish I could have talked to friends more than I managed to.
Anyway, if my papers next time are accepted I’m sure to be back. Thanks for everyone at OUGN for making this such a wonderful experience.