Top 60 Oracle Blogs

Recent comments

UKOUG 2011 Summary

This year's UKOUG 2011 conference in Birmingham was so packed and multi-dimensional for me that I decided it's going to take a few very different posts to describe my experiences without sending any unfortunate readers into a deep stupor, so the overall summary first and then a couple of more focussed posts to follow.


When I left the hotel in Edinburgh on Sunday morning and the snow was falling, I feared the worst after the horrible travel experiences almost exactly one year before but, in the end, a little de-icing and a small delay for incoming pandas weren't enough to hold me up too much. I had to walk round the block for a while when I got to the Hyatt to secure a smoking room but, with presentation preparation on my early agenda, it was worth it.

Which meant I only managed to register at the ICC in time for the final Performance & HA Panel of Oak Table Sunday. As I walked into the reception area where everyone was on a break, the welcoming sign of Oak Table Beer greeted me, courtesy of ... I'm not sure who - James Morle? Well, whoever was responsible, it was much appreciated - a nice BrewDog-produced lager. Lager with taste. Magnificent. So a quick one of those and another one to take into the panel session and we were away ...

Whilst it was good to see Dave Ensor coming out of retirement once again, most of the conversation (just as at the ACE Directors Briefing in Redwood Shores recently) was taken up by the inability of people to play around with Exadata without buying the damn thing first. Whilst I definitely sit on the side of the fence where Oracle are preventing passionate techies from learning about a technology properly (and will comment to Martin Bach's post here to that effect at some point), I felt a little sorry for Greg Rahn and Graham Wood from Oracle being confronted from all sides with questions about it. But it was all good fun in the end, before we retired to All Bar One for the annual UKOUG volunteers and staff thank-you drinks before the conference starts.

It was a little different this year because they hadn't secured the whole first floor, so it was a little fragmented and not as easy to mingle, but I managed to meet up with some old friends from Standard Life, Oracle regulars and even the odd Norwegian attendee I recognised from this year's boat trip. Despite sensible folk from Oracle trying to persuade me to eat, I still ended up going to bed without having had any food (and way too much drink, in that context). But at least I wasn't as late as usual and managed to avoid the very late night exertions in the Tap and Spile or Jury's bar!

I had a couple of presentations scheduled for Monday that were deserving of some final preparation.


I was still awake (or at least moving) too late to make Kyle Hailey's SQL Tuning presentation and decided not to attend Greg Rahn's Real Time SQL Monitoring presentation because I'd already seen him present on the same subject at Openworld. That is one of the benefits of attending multiple conferences, it makes it just a little easier to make agenda choices which are always horribly difficult at UKOUG. Instead, I managed to catch the second half of Connor McDonald's first presentation of the week on Partitioning. (This would be the first of many presentations I only saw part of, as I kept having conflicting meetings or would just get stuck in a coridoor talking to people ...) I always try to see Connor when I can because he is an amazing presenter so even if I know a subject well, I'm almost guaranteed he'll be entertaining and listenable and this was no exception. Good stuff.

Next up was Maria Colgan's presentation on upgrading to the 11g Optimiser and some of the strategies you can use to make the transition easier. Maria's another presenter I always try to see and although I think I saw most of her presentations at Openworld this year, I wanted to check out this one because I knew she'd be talking about SQL Plan Management and I'd be dealing with the same subject later in the day so wanted to check the presentations would be reasonably complimentary. She was her usual refreshing and honest self and between her and Connor, I was finally starting to wake up a little! ;-)

Next up was the first of the lunchtime OAK Talks sessions (where I did a 10-minute You Probably Don't Need Exadata talk) but I'll probably write a seperate post about those.

Back to my room, then, to remove a few slides from my presentation to bring it down from an hour to 45 minutes, get changed and start my usual pacing around, trying to calm the pre-presentation adrenaline rush a little and think about what I was going to say. I know from talking to other presenters that knowing you have presentations on a given day always takes up some of your energy and focus, but I suppose I'm among the worst I know - I'd rather have all my presentations out of the way on the first day if possible!

Eventually I was ready and made it back to the conference in time to catch the start of Graham Wood and John Beresniewicz talking about the DB Time Performance Method which had been one of my first choices when the agenda was published although, because we clashed, I did warn them I'd have to leave early to make sure that I was in Hall 1 in good time. It was a shame it was such a small attendance but I suspect that was largely because it was a *2-hour* slot at the end of the day so, as well as making it harder for people to choose (because it would clash with two other session slots), most people wouldn't have the stamina! Regardless, the bit I saw was their usual great stuff and I was disappointed I had to leave early.

When I got to Hall 1, it was getting towards the end of Alex Gorbachev's I/O Measurement and Planning session so I didn't see enough to say much about the presentation other than to say that Alex was perhaps a little more subdued than usual ;-)

Then it was time for me to present - another blog post will cover that - followed by Cary Millsap's presentation - definitely another post for that. Suffice to say for now that I'm glad (although unsurprised) it was so good because typically I never attend presentations immediately after my own because I'm too wired and tired to enjoy them but had to in this case to help Lisa Dobson charge her phone off my laptop. More to the point, it was now getting very late in the day and Cary didn't finish until 19:45!

There seemed to be general agreement that the day was far too long, but that didn't stop people heading off to the Exhibition drinks, then the Focus Pubs. It was particularly enjoyable having a good chat and some drinks with two of my three presentations out of the way!

The UKOUG directors know that Maria Colgan and I know each other so somehow I ended up being charged with making sure that she turned up to the Focus Pubs because she might be needed for the Inspiring Presenter Awards ;-) and she'd been talking about not bothering because of jet-lag! I kept having to leave other friends I was with to tour the hall to make sure she didn't sneak off! Once she'd picked up her well-deserved award (I vaguely remember being completely unsurprised at seeing Jonathan Lewis collecting a couple, too) I was ready for my bed, utterly bushed after what seemed the longest day and determined to enjoy a presentation-free Tuesday.


Tuesday began with one of those 'where is the presenter?' moments as a pretty packed Media Suite awaited the arrival of Thomas Presslie talking about his user experiences with Data Guard Fast-Start Failover. In the end, he did make it on time, as he ran past me into his room but then had the worst blow a presenter could have - his laptop containing all his demos wouldn't work so there was a small delay while he got his slides moved over to the in-room netbook. After that start, I was amazed how well he pulled it all back together, introduced whisky to his visual demo and was able to get across how convinced he is by DG as a low-cost high-availability solution. Good recovery (if you'll pardon the pun).

I was sorely tempted to go and see Connor again next but figured he's guaranteed a large audience and 11g ugrades are a subject I'm all too familiar with :-( so I plumped for Marco Gralike's presentation on "Drag, Drop and other Stuff. Using your Database as a File Server" XML presentation by way of moral support (I needn't have bothered - he had a pretty busy small room) and because I was more likely to learn something new. I did, but in areas I never expected to, including just how much detailed and potentially personal information modern image files contain! I loved his slides with the big, clear writing but it was a shame that it was very hard to see most of the damn screen in that room! Good stuff, though. I had yet-another-meeting scheduled so baled out half-way through as Mark Drake of Oracle started the demonstration (nothing personal, Mark!) ;-)

Where most people had a serious session-clash next, I'd seen several of the presentations at other conferences so was able to go and grab some food (I'd seen the conference lunch - cheese-central!) and then head over to the OAK Talks again before a pre-scheduled interview with Neil Jarvis from the UKOUG with Lavinia Foster on camera duties (something she only just managed to fulfill!). I think the idea here was to interview as many Oak Table types as possible for future use in the UKOUG's Oracle Scene magazine and, by filiming them, maybe release them as podcasts too but I'm not sure of the detailed plans. I must admit I was dreading this thing because it feels so much more intimate and focussed on me, me, me than a presentation but I think I just about got through it ok. I suppose I'll cringe if I ever see it though!!

The final presentation I attended was JB's talking about ASH Outliers. This was a must-see for me because it was a presentation all about a query JB's been working on to identify unusually long events from ASH data as a way of potentially diagnosing strange system hanging issues after the event. I'm going to do a seperate post on this presentation too because I think it's an important subject that warrants it and JB promised that I'd probably help him host the query as he is the *least* social media person I know. Or so he claims ;-)

As a sign of how long and tiring the days are, I was really concerned I was going to nod off a couple of times during this presentation despite how interested I am in the subject so decided I needed some rest before the party!

I liked the return of a slightly more formal party, with entertainers and gargoyles (!) and the masks. I must admit I thought the masks were going to be like paper ones out of a greetings card shop but they were pretty impressive and glittery and I noticed quite a few being tucked away as one of the better post-conference gifts for the kids. At least I assume they'll be for the kids. Despite everyone telling me I had loads of drinking time left, the memories of Bulgaria were all too fresh so, with a 9am presentation slot (one hour earlier than Bulgaria), I made my excuses and left. My word, I don't think I drunk past midnight once during this conference!


But, boy, was I glad I didn't as it meant I was early and prepared for my presentation on Statistics on Partitioned Objects. More in a seperate post, but I was reasonably happy with it.

I swithered over whether to attend too many presentations after that because, with all the early nights I felt I really hadn't spent enough time with all the friends I meet at UKOUG. It's actually becoming quite a challenge to see enough of people without being rude to others :-(

So I saw some of Latch Internals in 2011 with Andrey Nikolaev which was as deep as I expected and something I know I am much better suited to reading up on later, rather than try to focus on the subject mid-conference. I know it sounds strange, but some subjects and presentations just refuse to stick in my head.

I was also massively entertained by Connor's Optimiser Stats talk. The guy really is an astounding presenter and, as he said earlier, our two stats talks touched on completely different areas so although I knew the content, it was still as refreshing as always to hear his perspective. Top marks for producing a single slide animation sequence which, with Connor's perfect verbal delivery over the top, produced spontaneous applause from the audience. Not an easy thing to achieve at an Oracle conference!

Which just about left enough time for the final OAK Talks and then the last afternoon meet-up with a whole bunch of friends from Pythian and other places I've worked, Oracle, the Oak Table and, well, all those regular conference attendees that it's so much fun to meet up with. I loved this part of the conference, actually, as we all wound down and people had a last few drinks before heading off on their own particular journeys. It's a good job I had a pre-booked train reservation or I might have been tempted to stay much later to make up for my relative abstinence earlier in the week!

There will be more posts to come, as I keep saying, but I had a terrific conference, as I expected although I was hopelessly tired when I got home!

Thanks to all the people who put the work in to make this conference happen every year, to those who took the time to come to my presentations and to all the old and new friends that make it fun!

Usual disclaimer: My travel and accommodation expenses were covered
by the Oracle ACE Director
program. The Oak Table Network, or James Morle, or somebody, provided the splendid Oak Table Beer. Paul Vallee of Pythian stood his round and more at the bar as always but I still don't say enough nice things about him ...