UKOUG 2011 Conference OakTable Sunday by Alex Gorbachev

Dec 12, 2011 / By Alex Gorbachev

Tags: , , ,

This blog post covers day 0 of UKOUG 2011 — Sunday, 4th of December, 2011.

Since there were so many of us from Pythian at the conference, I’m adding my name in the blog post title. I think I will be doing it for all conference posts as I think I’ve been doing for some time. This year, there were ten Pythian folks attending UKOUG Conference and we did twelve sessions including multiple presentations, masterclass, RAC Attack workshop, round-table and 10 minutes OakTalk. I think it’s the record number of session Pythian folks did at a single UKOUG conference and the record number of Pythian peeps attending. A dozen of Pythian people in Europe and now even a sales guy in the UK mean that Pythian penetration in the UK database services business is close to the infliction point. This is ultimately a good news!

Most of Canadian Pythian representatives arrived on Sunday morning to London Heathrow. The flight was quite empty so some of us managed to get a good nap in comfort of three empty seats. Since AirCanada has power outlets in most long haul flights, I was planning to work on my slides all the way in as I usually do. However, this time I was sitting next to Christo and he kept me involved in the conversation and at some point I was getting sleepy and finally took a nap as well so I’ve done literally nothing on my slides. Oh well, at least I had some rest and it was good because I was up for almost 20 hours after we landed except a quick nap in the car from London Heathrow to Birmingham. By the way, if you travel two or more from Heathrow, hiring car transfer service makes more financial sense than train or coach and is also quite convenient.

Jury’s Inn has disappointed me again this year — refused to checkin us earlier and were not responsive to our prays of letting us in for a nap as we are tired Canadians that have just made it over the pond. Well, something has definitely changed at Jury’s couple years ago so this year I’ve got convinced that it’s been the last time I stayed there. Thanks you very much, your service s..cks.

Since Jury’s didn’t check us in, we went to grab a lunch in All Bar One where we occupied a table next to Mogens and the team — he was getting ready to deliver the first session at the OakTable Sunday. Where else to get ready but in All Bar One?!

After lunch, we tried to checkin again and were initially refused just 30 minutes before 2pm quoting checkin time of 2 pm. They finally, checked us in 15 minutes earlier. I suppose I should be grateful. At this point some of us decided to take a nap and others moved to the conference center to register and join the crowd at the OakTable Sunday sessions. The bad news was that I’ve missed most of the session but the good news was that I wasn’t hungry and could still attend the last session so I used my chance to attend one of the few presentations of the conference (of those few that I could attend). It was James Morle‘s presentation on using Oracle RAT (Real Application Testing) and Simora for capturing and replaying the database workloads. Very interesting topic and I learned quite a few details about RAT and James’ own product.
The tools are capturing traffic using different mechanisms. Oracle RAT is integrated in Oracle database and has special means to capture user requests inside the database instance and store them in binary representation. Simora captures user requests from 10046 traces. Because Simora basically creates workload scripts from the trace files, the scripts can also be customized later any way you like not just increase or decrease the rate which is what RAT can do. Both tools are facing the same major challenge — playback synchronisation — and they seem to address it in similar ways. With both tools, the playback is never the same even when you run the same workload twice and James went through multiple examples to illustrate it. The good news is that in most cases, we can use one or another playback mode to make the workload playback consistent enough in terms of applying the same level of stress to the system. The difference in the playback modes is how the tools are handling synchronisation of concurrent sessions during playback, i.e. how ordered your statements and transactions are. There are 3 levels in Oracle RAT and Simora adds another 2 methods but James mentioned that those two new methods are still somewhat experimental. One thing I noticed during the session is that I saw James presenting a bit more passionate in the past but I think it’s because it was the last presentation before the beer-time so his thoughts (or mine?) might have been outside the presentation area. :)

After that short but much needed break where we enjoyed some OakTable Network branded beer

we all moved to the OakTable panel on database performance and high availability. I was on the panel as well and I suspect I made it there simply because I still had my Pythian RED Mo going on. Here I am with Graham Wood posing before it was shaved off. This is one of the last photos I have with it still on my face:

Back to the panel… I’ve managed to capture a photo of the panelists where I’m missing, of course, since I was taking the photo:

On the photo above from left to right: Graham Wood, Dave Ensor, Jonathan Lewis, Greg Rahn, James Morle, Dan Norris (barely visible), and Cary Millsap. Imagine me sitting to the left from Graham and you can see the whole panel. Just make sure to imagine me like that:

The panel was interesting but half of the time we were preoccupied with the topic of why Oracle doesn’t let Oracle enthusiast access Exadata software and documentation as part of their training and research. It was clear that Oracle customers in the room were pretty much voting for having access to Exadata software even without original hardware but few Oracle folks seemed to be opposing it quite strongly. From my perspective, Oracle’s corporate position did introduce some impact there but what do I know?

After the panel few of us sacrificed our facial hair for Movember cause — Dan Norris, James Morle and yours truly publicly shaved their Mos off. James’ better half, Margherita, showed master skills operatic electrical shaver and it was a nice and smooth encounter of the electrical shaver for me — until that moment I never let these electrical devices touch my skin. It turned out to be not that dangerous but definitely needed some proper shaving after that!

By the way, while I have thanked everyone personally for their Movember donations, I want to thank you again here as many of you, blog readers, have donated either to me directly or to other Pythian Mos or to other Mo Bros and Sistas participating in Movember. It’s all very much appreciated. Thanks to your donations, I’ve personally raised $2,360 which includes Pythian’s $500 as the winner of the best Mo at Pythian. With Pythian also contributing $50 to each participant, we have raised more than $7,500 during our Movember campaign. I think that’s very impressive!

Of course, I had to get back to the hotel for a clear shave which I did and then joined the rest of the UKOUG volunteers crowd in All Bar One. The night carried on for quite some time — far far past midnight in Tap & Spile, the traditional UKOUG “drinkerie”. This past-3am-party has apparently set the pace for the rest of the conference so I’d not ended up in bed before 3am from that day on.

Monday was brutal but that’s the topic of another blog post… maybe I should skip that one and move to Tuesday?

Leave a Reply

  • (will not be published)

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>