“Miracle Open World” is the right name for the conference. There is no “Oracle” in it as I expected, but it definitely is about Oracle. I think it should be called Miracle Orange World:
Poll time now — Who is that in the picture? (I first thought I would ask what that formula means, but every reader of this blog is expected to know that already.)
The conference opened yesterday. You probably won’t be surprised to find out that it was opened by Mogens Norgaard. He started by disclosing the difference between other Oracle conferences. While normally you would expect 80% of the conference to presentations (or “useless stuff” to quote him) and 20% for social networking. The Miracle event leaves the rate of “useless stuff” at 20% and raises the social networking component to 80%, and last night I learned the secret how to do that.
Mogens spoke about some other “useless stuff” like how unemployed DBAs could easily find a job during the conference and so on. Mr. Miracle Iceland tried to give as an idea of how Miracle Iceland operates, and it looks like a dream job — 96% of the population of Iceland (according to Mogens) work at Miracle, nobody really works (they don’t even have managers, sales, and so on), and customers (I expect the rest of population [4%]) are happy. What a dream job! But I wonder if they have any salary. At some point, we had a short break after this “useless speech” and switched to the other 80%. Thanks to the stuff on the tables, it was easy to do!
The next session was meant to be about High Availability, delivered by Michael Moeller, but Cary Milsap opened it with FUN (Functioning Under Normality). Cary talked about a lot of other “useless stuff”… “gold turtle” is the only thing that I remember. Michael was great as usual, but unfortunately we couldn’t get to the last slide due to the number of unfortunate events which provided a live demonstration of what can happen to your high-availability environment. The lesson is: no matter now many measures you take to prevent a disaster, there is always the time when they all fail. This was exactly the situation — all 42 backup measures for his presentation failed right in front of us.
At the end we were presented with a poll and needed to provide 42 answers to a single question: name 42 reserved words from Oracle SQL that would fail in the statement
CREATE TABLE (N NUMBER);. This wasnâ€™t easy. I can definitely name 42 reserved words but I know that some of them can actually be used in a
CREATE TABLE statement.
We quickly filled in the questionnaire and hurried off to the Party House (room number L142, what a surprise!). There, the second 80% part of the event was successfully completed. Needless to say, it was quite late and usual suspects left the house at… well it was still dark, so relatively early.
This morning I was able to wake up early and felt surprisingly well. I think Danish beer is designed to be hangover-free. I finally glanced at the agenda and discovered that sessions were 90 minutes. This is an excellent slot size. I went to Tuomas Pystynen presentation (again) and I could see the difference — it was much better than the same material delivered in a 40-minute slot at SIOUG even though the slides were the same. Tuomas was able to go through it slowly and discuss the questions from the audience without any hurry. It was much easier to the attendees to get into the material and have some time left to think about it.
For my next session, I choose Anjo Kolk’s speech on never-learning developers. Yes, the speech — he didn’t use any slides as the concepts he was promoting didn’t require any visual support. I think we should actually prohibit projectors in the morning — they are too bright and hurt the eyes in the first half of the day. But I digress.
DBA stands for “Default Blame Acceptor,” according to Anjo. He made the point that a DBA should never take any investigative actions at a developer’s request. A DBA should ignore them as long as possible and not start looking at the database stats and whatnot. Instead, a DBA should go with the developer to his desk, let the developer show everything, and go from there. The fundamental difference is that, with Anjo’s approach, the problem and solution will look like it’s coming from the developer himself with his code, and the DBA just sat there playing an almost passive role. Contrast with the DBA investigating that slow, stupid database, and then asking the developer to change his fine code to work around the problem of that crappy database that the DBA cannot run well.
Another interesting idea from Anjo is that instrumentation in Oracle should not stop at
DBMS_APPLICATION_INFO.SET_ACTION, but extend to
V$SESSION_LONGOPS being filled with progress info on long-running jobs. I must remember to recommend that in addition to normal instrumentation to all developers.
The after-lunch session for me was the future of performance optimization by… well, here’s another poll: Who do you think the speaker was? If it takes you longer than a few second to answer, or if you answered incorrectly, you definitely missed the Oracle performance tuning trend of recent years. In fact, Cary proclaimed that the term “tuning” will be deprecated and considered bad practice. Instead, the focus should be optimization. I also very much liked the “Why Guess When You Can Know?” part — very relevant to BAAG party ideas.
The last session of the day was a hard choice, but I finally managed to stop at “Oracle Change Notification” by Morten Egan, and I should say that it was absolutely a must-see. The presentation started without Morten himself because his computer was doing the talking and presenting. Morten was updating a table behind the scenes, and his computer was receiving notifications triggering the next phrase. During his presentation, he conducted a demo, and a special PL/SQL procedure was reading the notifications to us. It was an excellent approach to demonstrating asynchronous actions.
The Gala Dinner and the Beach Party are tonight, followed by the Party House, of course. As usual, keep an eye on your RSS reader — more details should come tomorrow!
Thanks for this, mate. It really gave me a taste of the event (although not quite like the real thing). I agree about 90 minutes, but most wouldn’t. (Actually, I prefer 4 days, but that’s another story.)
Please keep up the good work, even if it becomes more difficult as the social events kick in!
[…] different: Alex Gorbachev of Pythian is having lots of fun at Miracle Open World conference, and he writes about it in the Pythian blog. Boy, does it sound like a great conference. Posted in development, musing, nfs […]
[…] have already mentioned the 80/80 percent rule for the MOW, and I bet you suspect that 42% of “80% of useful stuff” was done after hours. And […]