Doug's Oracle Blog

Well, that was a nice start to the day! Someone came up to me just before Wolfgang Breitling's presentation to point out that something from my presentation had helped him fix a problem at work last night. It was the OEM Raw Data drill-down that shows you the underlying ASH data for sessions, including backgrounds. He'd used it to identify the timed event leading up to a problem with a crashing smon process. Actually, the more people I speak to, the more I get over the presentation. Most people recognised I made the best of a bad situation but I'm glad that someone actually learned something, too!

Wolfgang's "Anatomy of a SQL Tuning Session" was one that I managed to miss at UKOUG and I'm glad I made it this time. It revolved around taking a single SQL statement that took 9 minutes to execute and walking through various tuning iterations, using modern tools and optimiser possibilities, leading to a sub-second execution time. It was a natural extension of Wolfgang's Tuning by Cardinality Feedback that I've often recommended to colleagues. He covered a variety of techniques including :-

- Converting parts of the statement to scalar sub-queries
- Subquery factoring
- Transitive closure, with a useful tip that it may be worth experimenting with specifying apparently redundant join predicates to give the optimiser more information to work with.
- Using the Outline part of 10g DBMS_XPLAN to identify the set of hints that would create a specific plan and then using some of them yourself (but this is far easier if you name your query blocks).

What I particularly liked about this presentation was the way that Wolfgang illustrated execution plan steps and changes with some nice slides, highlighting a few steps at a time. But it made me feel a lot better about my mobile phone going off the previous day when Wolfgang's *own* mobile went off during this one

Next up was Neil Gunther with "How to Quantify Oracle Scalability - Part 1", a presentation about applying his Universal Scalability Law (USL) to Oracle systems. There were quite a few high-level points I picked up from this.

People often assume that the purpose of a model is purely to predict the future but it's probably just as (or more) useful as a method of validating test results because in his view 'Data comes from the devil and models come from god'. I'm not sure I agree with that. Test results, even incorrect results, represent reality to me that can't just be explained away by Maths which doesn't agree with them but I suspect that's my peculiar perspective. I much preferred the suggestion that we use models and data together because, whilst he might not trust my data, maybe I don't trust his model yet?

He talked about how the USL allows for the phenomenon of reduced throughput as workload increases which is something I think I've seen before by adding Coherency to Amdahl's Law. The USL is definitely worth more investigation. As a non-mathematician, though, I suspect I always struggle with this stuff.

Which was why I was *so* relieved that I made the tough decision to skip Riyaj's presentation and stayed for the second part of this two-hander by Peter Stadler - "How to Quantify Oracle Scalability - Part 2". This was a more practical examination of the USL in relation to Oracle systems and as someone who is very interested in performance in general and the relationship between Response Time and Throughput in particular, this hit the spot. What was slightly bizarre, though was when I recognised the URL for this blog post and the test results come up on screen. I think I'm right to say that this is the second consecutive Hotsos Symposium where this one post has been discussed (by Cary Millsap last year) so I must be doing something right

Peter spent the next 20 minutes or so talking about some of the comments on the post and plugging the results into the USL. I must admit to being slightly surprised by the fact that Peter didn't think to drop me a mail to let me know he was going to talk about it so much because I might have missed it and it was fascinating! He talked about the lack of detail in the results, but that was because the blog post had an extremely simple message - are you looking for High Throughput, Low Response Times or both? Regardless, if he'd asked me, I could have given him some more information to work with. For example, there was some discussion about measurement errors in data and performing multiple runs to address that which is something I did, but only published one set of fairly representative results.

So it was all a bit strange and unexpected, but utterly fascinating to see someone apply a mathematical approach to my empirical results. I hope that Peter might post the slides and add a URL to the blog post so that everyone can share what he found. I think that's the point of the comments thread and of blogging in general - sharing information and knowledge and building a discussion.

Next I managed to eat a little bit of much-needed lunch with Paul Matuszyk and then had to get ready for my important work call.

You managed to capture a couple of minutes of my presentation when there was a picture on the screen!

I really don't know how to blog about this because every time I feel I'm honestly self-critical, everyone thinks I'm close to suicide or something. I like to think I notice both the good and the bad but am probably more likely to speak openly about the bad. If you ask me it's a Scottish thing about not getting above yourself or blowing your own trumpet too loudly. Or maybe it's just that the only way you can improve is by noticing the bad stuff and fixing it? Someone once said to me "You can't be an insecure overachiever without first being insecure" and they aren't Scottish, so I shouldn't generalise. Updated later - it was Cary Millsap who came up with this line. I loved it when he mentioned it a few months ago and still do.

In this case I'll try to be even-handed and finish with the positives!

Negatives

The whole point of this presentation is that it's about 50 minutes of demonstrations and about 10 of slides. For the first 30 minutes, the demonstrations would not work. If you think that's a success, there's something the matter with you and you should probably never give any presentations!

This is the killer. In retrospect I know what the problem was and have fixed it previously with the help of others but did not do the same on the new laptop! As I moved in and out of wireless range, the lack of a network connection at the Windows end completely threw VMWare and my demos. But, believe me, when your demos have been working for a couple of weeks, you don't have long until your presentation and they stop working intermittently, it's difficult to be cool and analytical. I was so distracted by other things that might go wrong, I missed one and, having missed it, I wasn't cool enough to recognise the symptoms. Fortunately, an old Hotsos friend in the room came up with the goods in the form of a little Sprint wireless box that got me hooked up. I kissed him, but there were no tongues involved.

With only 30 minutes left, there was very little I could show and it completely ruined the whole flow of the presentation, which I'd worked so hard to get right and which I know can be terrific, because I've done similar presentations before and had been looking forward to doing the best version yet. People took the time to come and see it, I let them down and I'm sorry about that. Nobody is likely to change my view on that.

This presentation is a big deal to me, never mind anyone else. I work hard on these things, try to cover all the angles, take time off work and all because I like teaching people new stuff. When I take a week off to attend a conference, but also to present, it's a bit of a blow when your one shot fails. If you're not passionate about your presentations (and I somehow doubt anyone isn't) again, you shouldn't be presenting.

Positives

Sh*t happens when you get involved with computers. Yes, folks, I know that, but that's also why you play around with demos for a long time to minimise the possibility that it will!

This might not sound like a positive and it's dangerously close to sounding like an excuse, but the fact is that I've had 2 out of 3 presentations go very wrong recently. One of the consistent factors in this is that I switched to VMWare because I had to to run 11gR2 on Windows. It's not the same as blaming VMWare to say that I've been having to deal with stuff I haven't for a long time. When everyone was talking about VMWare as the way to go for presentations, I remember thinking 'I don't know, just seems like more moving parts that might break to me.' and I kept reading blog posts about demos being broken and then fixed just in time, all of which were on some virtualisation platform or other, but didn't have the confidence to say something. All I know is, say what you like about Windows, but I've hardly ever had a problem in multiple 2-day course teaches of performance and OEM stuff! Still, it's down to me to get on top of what are some simple issues.

Listen, I know I can present. No false modesty round here. So, to wrap this up on a positive note, I'm well aware that there are few of the people who I see present who could have managed to get through that first half hour, make people laugh, keep thinking about the problem and manage to get a short demo of Swingbench into the bargain. I think I also managed to salvage something out of the last half of the presentation without completely collapsing into a heap. Sure, I was a bit brain-addled by then, but I would have liked to have seen how others might have coped Frankly, I kept waiting for the room to empty (I've seen it happen) but the vast majority stuck around to the end. Maybe they were sadists! LOL

I am not and will never manage to be happy with that presentation but lessons have been learned, it's just a presentation and there'll be lots of others. I know that.

Postive Solution 1 - If anyone wants to try to grab me while I'm at the conference, I'll show you the screens and demos. They're very cool

Positive Solution 2 - I think Alex Gorbachev might try to arrange for me to repeat the presentation properly as a webinar. I've already done this at my current customer site once (on the 10g stuff) and it went reasonably well. I'd also ... get this ... asked Marco to video it for me yesterday with his snazzy mini-setup because I thought I might post a few bits online if they were particularly good so those that can't make it to conferences could get a taste of it. Actually, it appeals to my cold, self-deprecating sense of humour to post some of it online soon. I promise I won't make it too self-flaggelating though!

P.S. For Paul Vallee. Paul it is not all good

A 3:30 start, which gave me lots of time to work on my demos and by breakfast time at 7:30, I felt in a reasonable place (but not quite done yet) and headed down to eat ... wait for it ... some fruit! Then again, I'd eaten so little the previous day that I had to eat something.

First up was Gary Goodman opening the event and introducing Tom Kyte's keynote with :-

a) A video of Tom and Hotsos regular Patty, Disco Dancing to mark tomorrow night's Disco theme. Personally, I thought it was hilarious and I hope it gets posted online somewhere. I'm not sure everyone found it as funny, but it did the trick for me.

b) A gift of a custom set of poker chips because I believe Tom's a player (of poker, that is). Tom's keynote was about (looks for title in notes and can't find it) the mistakes we make *because* of our experience and our assumptions. It was thoroughly entertaining and kicked off with a quiz for four volunteer participants who had to answer 12 apparently simple questions very quickly. They were trick questions, of course Rather than me taking voluminous notes, I'm hoping that Tom might post the questions on his blog (hint, hint) even though they were meant to be answered quickly based on verbal questioning.

Oh, and I couldn't believe it when my phone started ringing, for someone who's normally exceptionally careful about that stuff. I had it on silent last night, missed a text from Alex G and so had taken it off silent. Clown! (In retrospect, this should have been a warning sign.)

There were tons of great examples of where clever people have gone spectacularly wrong through over-confidence. We're often wrong when we answer quickly based on our prior experience because things change. I suspect I sometimes frustrate colleagues by not giving the snappy answers they expect, but I know that Oracle stuff is often not as obvious as it seems at first sight. There was even a sighting of Martin Widlake's "Making Things Better Makes Things Worse"  but no name-check, Martin, so you're only *almost* famous

I particularly enjoyed the video I hadn't seen of Richard Feynmann talking about the uncomfortable state of confusion and feeling stupid. (With hindsight, why did I not see the warning signs ...)

Then I skipped two sessions I was looking forward to but, as I've said before, I need to take care of my own presenting business before I can enjoy and learn from other people's presentations. However, I bumped into a friend after Richard Foote's session and he thought it was amazing. My friends first trip to Hotsos is going well.

Sitting in my room, I was really happy with how the demos were looking, raring to go and went to iron my shirt. Damn, the iron wasn't working and I really didn't have time to wait for another to show up. No problem, though, because I always carry a special shirt that *really* doesn't need ironing - just hang it up in the bathroom with the shower on. (Another warning sign, though?)

I'm very superstitious about using any new kit for the first time, so I was still nervous about a couple of things with the demos

1) Performance, because I know I'd run them all weekend and have a new powerful laptop (more on that another time), but the demos are designed to hammer the machine. So I thought I'd disable everything I didn't need running.

2) Driving the projector because I'd never driven one with the new laptop. I arranged to check that my laptop would drive the projector ok. It seemed to, but one of the demos was behaving strangely. No time to investigate because Dan Norris needed the podium, so I decanted to Alex's Battle Against Any Guess presentation because we'd been discussing it, I know he was concerned about how it would go and wanted to show a little moral support.

In the event, he needn't have worried. Personally, I find his Russian accent all but impenetrable, particularly after we've had a few beers and are both shouting at each other, but I'm sure the feeling's mutual! But the fact that he had the room rolling about with laughter at regular intervals meant it had gone very well! I hope he still does plenty of technical presentations with lots of demos, but he proved he can do something a little more conceptual and message-driven. Better still, I'd been frantically trying to get my demos working at the back of the hall by re-enabling services I'd disabled and they started to work I turned round to Cary Millsap who was hiding their too and dramatically mimed wiping the sweat away from my brow. Time to head back into the other room and watch the tail-end of Dan Norris' Database Machine presentation. I didn't see nearly enough and was completely pre-occupied by what was to come, so will need to catch up on the slides later.

Some interesting Q & A at the end which were slightly less interesting for me because my demos started playing up again! So, a room full of people, I'm rebooting my laptop, nothing is working and then, bang, it's time!

Whilst my presentation was originally part of this post, my review is so long-winded and personal, I've moved it to another post that people can choose to skip over! That way it doesn't detract from the good presentations I saw yesterday, such as Kevin Closson's "Ten Years After Y2K And We Still "Party Like It's 1999"".

Excellent, as always, and packed full of information, it also sort of tied up the way the day opened, albeit in a very different tone and style that things change. Kevin has the perspective of having been in the industry a long time, so he's seen the past, but also being able to see where we are now and where we're going, so he concentrated on a few of his pet subjects that I'm pretty sure you'll see cropping up on his blog in the near future but, to give you a taster, he talked about the merits of SMT on Nehalem EP chips for different workloads, which is something we'd been talking about at the opening reception. I'm pretty sure I can't disable it on my laptop to give it a try, but I might have a word with him about it anyway. He talked about Flash, Direct-attached storage, NFS, virtualisation and all that good stuff as always but because I missed the first 10 min for a much-needed break and couldn't read the slides because I forgot my glasses, I'm going to have to try to hit on him for a copy of his slides later! Oh, and I was wondering why the presentation was well attended but not packed. I'd forgotten Tom Kyte was speaking next door. I was happy with my choice.

I did have one more presentation scheduled, to see the first of Kerry Osborne's but I was pretty low at this stage and sleep deprivation was kicking in, so I headed back to my room to catch up with some blogging and tweeting (thanks for the support) and then it was time for dinner with the Oak Table crew. Very nice it was, too, hats off to Marco and Carol for choosing that place, the company of all and for Carol's consistently top-notch organisation and geek-ferrying. Nice big lump of beautiful steak, some chips and a beer. That's more like it! A few more drinks back at the bar but I made my excuses and left at 11:30 despite some vain attempts at hypnotising me into staying!

Oh, but the weather was still rubbish all day

Regardless of everything that happened yesterday, I now have 3 days of no responsibility and can just learn, socialise, eat and perhaps sleep a bit more. I woke up at 3:30 again this morning, but turned over and managed another 2 hours. Bliss!

Hard at work preparing for my presentation yesterday until it got nearer to 5pm and people started coming in on their flights. First up was Kevin Closson, who looked a somewhat sad sight (by his usual standards), clutching a bag from a PC store containing what he hoped would be the crucial bit of software that would help him solve his over-written MBR. I waited for him in the bar and discovered Mark Bobac had Ubuntu Live on a USB stick. I suppose I could have gone and told Kevin, but I knew he'd be down imminently

It was also a brilliant experience to bump into Paul Matuszyk again, who I worked with briefly 4 years ago at Sky. This is his first conference, largely inspired by a set of Symposium 2006 notes I gave him and my encouragement so it was a) great to see him again and b) pleasing when I saw him coming out of Richard Foote's presentation, blown away by how good it was. I think he's already decided it was a good decision

Then it was the usual suspects, the usual drinking and some chat with KC, who always has smart stuff to say. In amongst that, I liked him pointing out that my new iCore 7 laptop is probably what they used to euphemistically call a 'Portable' computer. He's not wrong there

Not too late to bed and really looking forward to the first day of the conference and my presentation.