UKOUG Begins

I have a feeling I might be blogging a little less than usual this week.

Work has been busy and my laptop's power supply packed up late last week so that I didn't have a chance to get a replacement in time for my presentation. In the end, I gave in and bought a new laptop. The make and model will be kept to myself, lest I be criticised for buying something over-priced and under-specced. Let's face it; once you find yourself shopping in PC World in a hurry, you know you're getting stiffed anyway At least I can relax now that I'm suitably equipped again. Once the laptop was fully configured, though, I had lost loads of presentation prep time so I'm going to be catching up a bit this week. Thank goodness I don't have to present until Thursday!

I arrived at about 5pm yesterday after what seemed an absurdly short pre-conference trip, and less eventful than Marco's! As soon as I walked into the lobby of the hotel, I saw someone in a small bunch of people waving to me. My eyesight is getting worse these days, so I couldn't really see who it was, but thought it safe to wave back It turned out to be Tom Kyte, sitting in a group that included The Official BAAG Party delegation, Marco and Melanie Caffrey, to name a few. Kurt showed up later as well as Lisa Dobson and Dan Fink, but I'd better stop in case I miss someone. I was musing on where we'd all been a year ago. Tom didn't have a beard, but I must admit it's growing on me (as well as him, naturally) and Kurt was getting ready for his first big conference presentation. If I'm not mistaken, Alex had three to prepare for! Marco is the first-timer this year It is a pleasure to be able to meet up with people again, shoot the breeze and see how life's changed. Everyone seemed on top form. (Oh, but I'd better hope that Connor MacDonald doesn't find out that people may have discovered I'm not the only member of the community who lives with a cuddly toy or two! I reckoned I was safe as he's not coming over for this year's conference. Speaking of cuddly toys, there are some new additions ....)




Thanks to Lisa for the Penguin (who has no hole, despite looking like a glove-puppet) and Alex for the Sens Moose. You can tell how excited Geoffrey is about making the trip but Brian the Footballer is looking forward to meeting Andy C and finding out if he really believes he can be the next Paul Scholes.

When it came to the Indian vs. Burger and Chips decision, I was only ever going to go one way. (Three Belgians, a Dutchman and a Scot. Mmmm, do you think there might be a 'Chippy' thing going on?). The food was nothing special to be honest but the conversation was great. I was a little tired after working on a Production problem early on Sunday morning, so it was a welcome early night in the end.

Registration was very straightforward this morning. Then it was along to the speaker lounge to pick up the speaker gift.



A clockwork radio (and torch) appropriately.

I seem to be bumping into people I know constantly. That's something that's changed a lot over the years ... I used to be able to traipse around the conference largely unnoticed but I think I prefer it this way. It's one of the enjoyable aspects of UKOUG. As Marco pointed out in his blog, it feels cosy, even if there are about 2,900 people.

I skipped the Oracle sales-pitch and went to Hall 1 in time to hear the Tom Kyte bit, discussing improvements to Oracle over thirty years and focussing on 11g improvements. He had a terrific picture of an old 1Gb hard disk drive, compared to a 1GB SD card. I'm not sure I've seen it before. If not, maybe he'll post it on his blog for all to see. Good presentation, but I suppose you guessed that already.

Next up was yet another CPU presentation - Matt Penny's "Applying Critical Patch Updates - what's worked for us and what hasn't". I'm attending a lot of these at the moment as part of an exercise at work to get to grips with this minor nightmare. In this case, it was interesting to hear a real world DBAs perspective although it was probably a little light on technical detail. He focussed on the importance of having a consolidated patching 'cookbook' or process documentation because there are so many things to think about apart from the database when you're taking regular outages on production systems. I've no idea why it was a 60 minute slot, though, given that he finished in 45. I would rather have seen the extra 15 minutes going to the next presentation I attended.

Having logged in to work to reply to a bunch of emails about a project that's going live this week, I ended up being a little late for Jonathan Lewis' first presentation this week - "Playing Russian Roulette with Silver Bullets" but I think I got the general idea Although there was quality technical information, as always, this presentation focussed on why Silver Bullets are a bad idea and discussed the human issues as well as the technical. Just picking out a few highlights ...

He talked about the three W's

• What is the problem? (Because you need to understand it properly first.)
• Why will this fix it? (If you don't know why a fix works, how can you justify it or know whether it really will fix the problem.)
• Where will I pay the cost? (Probably the main point - always question what the down-side will be elsewhere, because there will probably be one.)

For a specific example, he went on to one of his pet topics - the over-use of optimizer_index_cost_adj, which he described as a hack to help the optimiser if you're running Oracle 8i that shouldn't be used in 9i and definitely 10g, where we have the preferred option of using System Statistics. He showed an example I've seen before where, because low index costs are rounded down, they both become 1, so Oracle has no sensible way of choosing which is the 'best' index and just picks the one that's first in alphabetical order.

In fact he suggested that if you've got a really 'good' index, call it AARDVARK He specifically told us not to quote him on that, but I'm not to be trusted!

He rounded off the presentation by suggesting that a Function-Based Index could earn the label a successful silver bullet, if done properly! The reason being that it's a solution that has a very narrow scope, designed carefully to fix a very narrow problem, which reinforced one of the main messages of the entire presentation. Silver Bullets are dangerous because they tend to have global scope when you're only really interested in fixing a local problem.

Now it's time for some presentation prep. I've resigned myself to missing lots of interesting presentations (and the agenda looks terrific again) because, ultimately, if my own presentation isn't right, it'll ruin the week for me so I'd better get that sorted first. Besides, I have a Blogger's Meetup this evening. I'll have the two important tasks of helping to uphold the British reputation for working our way through a bar tab and of escorting a special guest or two