Daily Archives Wednesday, May 2008

AppsLab FAQ: What if Someone Posts Porn?

I’ve decided to start a periodic feature I call AppsLab FAQ, which serves several purposes.
First, it gives me topics to blog when I’m short on ideas, which happens frequently.
Second, it gives me a permanent place to store experiences and lessons learned in a given subject area. So, the next time I get asked the question, [...]

A Fix for Check Constraints That Harm Cardinalities

I wrote a while ago about the harmful effect of check constraints on query optimization. I’ll pause while you catch up on that …
Anyway, this appears to have been addressed in 10.2.0.4, and it’s now safe to go back into the water on applying check constraints. Bug 5891471.
The bug description also notes that you can [...]

Orcan/OUGF Day 1

I’m typing this entry as I wait for Moans Longballs Nogood to kick off the second day of the conference for the Swedish contingent, which is the first day for the Finnish contingent. Here’s how it works. Both user groups co-operated in organising the event

Day 1    Start at Stockholm and first day of Swedish event. Sail overnight to …
Day 2    Helsinki. The Finnish contingent board the boat and we have one day of both user groups together. At 5 we’ll sail back to …
Day 3    Stockholm. The Swedish contingent depart and the Finnish contingent have day two of their conference. At 5 they sail back to …
Day 4    Helsinki, where the Finns depart.

OK, better stop for now as Moans is about to start speaking. There has been a mix-up and the Finnish agenda shows "We Have Probably Never Tuned Correctly" whilst the Swedish agenda shows "We Probably Need RAC". It’ll be interesting to see how he combines them. I’m sure he’ll manage it in his own inimitable style ;-)

[Break]

In the end he spent at least 45 minutes talking about why we Probably don’t tune right (as well as raising a few laughs, as usual). He talked in particular about three things we do wrong.

We talk about Best Practices. A lot. He highlighted that around 42% of the presentations at last years Openworld had ‘Best Practices’ in the title. In many cases, these are for new products so, without great experience of product use, they resort telling you the obvious (very similar to what I’ve seen for Audit Vault!). He reiterated a point that Cary Millsap has made in the past - maybe Worst Practices would be a better idea? We could stop doing things we *know* to be a bad idea!

We try to use Counters to measure performance, for example the number of waits when we don’t have their duration because timed_statistics=false

One his old favourites - "Guess and Grimacing" was the final point. He made this even funnier than I’ve heard before by focussing on how many more people are involved in the guessing process these days because of the modern balance between the number of IT people and bureaucratic surroundings (or Project Managers). He proposed this might be a solution for global warning - reduce CO2 and heat generation by reducing the number of managers. More boots on the ground vs. Majors, which I was discussing with M2 and Mark Rittman last night and is very timely. I’ll say no more than that ;-) He highlighted that awful moment when you’re in a room and someone says - "Let’s get all the possibilities on the white board". Moans has a way of cutting to the core of the pain and absurdity that crops up in organisations.

He went into a funny diatribe proclaiming that men (and he was specific about gender) should be "Allowed to work" instead of "Reporting". In Mogens view (and I have a lot of sympathy with this, particularly recently) work is the most important thing to a man because it builds his sense of worth and achievement. If you stop him working towards an achievement he can value, his testicles become smaller, hormone production slows and eventually his testicles fall off!

Towards the end (or at least as I popped out for a smoke!) he summarised most performance optimisation methods as such, which is essentially what I was saying yesterday ;-)

  • Trace the actions?
  • How much contribution to the overall time do individual actions take?
  • Reduce those that contribute most.
  • Repeat

After that, I popped to my room for a much needed nap which meant I missed lunch, plus one whole session and the start of Daniel Fink’s. I was aware of some of Dan’s thoughts on tuning from his blog anyway, but I like hearing him speak and he has so many good points to make. It really isn’t about ‘Tuning SQL statements’ it’s about understanding your business; your applications; your systems. But to say any more wouldn’t do it a justice, particularly as I was late! You can read more of Dan’s thoughts over on his blog.

Now it’s time for me to start worrying about my own presentation again, but at least I’ve had enough sleep this time!

UK Higher Education Seminar : Wed 25th June, 2008, Reading

If you’re in the UK Higher Education Sector and are users of Oracle’s business intelligence software, one of the things you’re probably particularly looking forward to is the inclusion of Oracle Business Intelligence Enterprise Edition in the “Campus License” agreement. We’re particularly pleased therefore to be putting together a one-day seminar towards the end of [...]

Comcast Buys Plaxo

LewisC’s An Expert’s Guide To Oracle Technology

Have you heard of Plaxo? I’ve been a member for a while though I’m not a heavy user. Plaxo is sort of a social networking address book. It’s a decent little site.

Have you heard of Comcast? They’re sort of the <a href=”http://www.google.com/search?q=comcast+evil&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en

Orcan 20th Anniversary Conference - Day 1

Well *that* was an interesting 24 hours (for me at least - it might not seem so interesting once I write it down …)

After a whole two hours sleep, I had to get ready for my taxi to the airport. The trip itself was largely uneventful but the flights were so short (with a couple of hours wait in Copenhagen airport in the middle) that I didn’t manage to catch up on any sleep, as I’d planned. Still, the train from Arlanda to Stockholm was fast and easy to find and a short taxi trip found me at the cruise terminal. (I’m glad I took the point-at-piece-of-paper approach with a taxi driver, though, because it sounds like Mark Rittman conducted a solo walking tour of Stockholm’s harbours!)

I was the last person to arrive so fortunately Patrik Norlander was waiting for me with my room key, breakfast tickets, bag and T-shirt. (Some nice goodies in the bag, too.) Amazingly, I managed to find my way on to the boat and into the dedicated conference area. I knew I’d found the right place when I saw Jonathan Lewis and Julian Dyke having a coffee and a chat so after catching up with them, I thought I’d better find my room and get changed. Rooms with a window are at a premium but this particular vessel has rooms with a view not just on the traditional outside of the boat, but on the inside, too, overlooking the central promenade of shops, bars and restaraunts. I’m sure most would prefer the sea view, but I can think of at least two stowaways who would much rather spend the three days people-watching. It was very decent of Patrik to think of that.

I only managed to see one presentation - Jonathan Lewis’ "Optimising Through Understanding" which was an updated version of the web seminar he did for Oracle. I couldn’t possibly do it justice via a blog entry so I suggest you go and look it up on his website, but I suppose one of the main themes was the use of Query Block identifiers when writing queries for 10g and above to help document your work and aid analysis of the subsequent execution plans.

During the next slot I had one more run through the outline of my presentation, checked the demos looked ok and then it was my turn.

Departure from Stockholm coincided with the start time so there I was, slightly frazzled due to lack of sleep, ready to start talking and I could feel the boat start to move. Hardly bobbing along, but unsettling nonetheless. Based on the audience reaction and a few later comments, I think it went fairly well but I must admit it’s one of those presentations that I know I’m going to get just right one day and I haven’t yet. There’s so much I could show and I don’t think I’ve made the right choices yet to make the most of an hour. When I finished presenting, Patrik presented me with an unexpected speaker gift of an iPod shuffle, which was a lovely gesture and much appreciated. The Boys have been after one for *ages* but I’m not sure how they’re going to share it around 28 of them! Maybe they can take turns or something.

By now it was after 6 so I went for ‘just one pint’ with Mark Rittman and Michael Moller. Which, of course, turned into more than one drink and reached the stage where I no longer had time for a quick nap. You can imagine the state I was in by now! I *almost* bailed out of the evening dinner but I thought that would be pretty rude so I made it through that on the last dregs of adrenalin, a nice dinner, some alcoholic fuel and some fascinating chat with a couple of guys who are over from Estonia. One of them suggested I should try to speak at an Estonian event because he’d enjoyed the presentation. Better still, maybe Oracle will schedule the full two day seminar there? Just an idea ;-) In the end, with extra assistance in finishing a couple of the courses (thanks M2) I made it through to my head hitting the pillow at 11. Which meant I woke up at 4am because 5 hours sleep is about as much as I manage most of the time -( It’s probably the first time I can recall waking up with a wallet full of free drinks vouchers which I’d found no use for the previous night. I’ll do my best to make up for it this evening, though.

There were some interesting sights on the boat between 4am and breakfast to keep me entertained. I realised that some people were still awake and drinking and very, very drunk! People obviously have a great time on these trips ;-) In fact a few hours later I went to the bathroom (bear with me, there’s a point to this) and someone’s jacket was lying on the floor, which seemed a bit weird. Not quite as weird as the sight when I turned the corner and an older gentlemen was propped up against one of the sinks, half sitting in it, fast asleep! I thought I’d try to make a quiet exit and leave him to it, but I think I might have made too much noise because I saw him open his eyes as I left. He looked pretty confused. With hindsight, I wish I’d taken a photo!

Then when I went towards the buffet restaurant for breakfast there were two main categories of people. The older generation who were up too early and keen to make a determined assault on the buffet before the crowds arrived (that’ll be me then) and the younger generation who were dressed in a manner that suggested there’d been no sleep at all or at least not until *after* they’d had their breakfast. The boat’s an fascinating experience with an interesting crowd. I feel more at home here than at most conferences - it’s like a Saturday night out in provincial Britain but without the violence.

Anyway, I’ve been up for 4 or 5 hours, so it’s 9:30 local time (switched time-zones again). We’ll arrive in Helsinki soon but I might go back to bed for 40 winks. I’m tired again already ;-) I didn’t have my wireless access sorted out yesterday but, now I have, expect more later. So far so good, though!

U Can Has Beta Invites

Last week, I alluded to Twitter as a place to get and barter for invitations to beta services. Well, blogs are often a good place for that too.
I have invitations to several services that may interest you, gentle reader. I don’t want them to go to waste. And since everything has a social network now, [...]