Yesterday was the developer training/seminar that I mentioned in an earlier blog. It was okay I suppose but I made a bad decision to drop analytics because I didn’t feel I could do them justice in the available time. Then I finished half an hour early 🙁 Based on verbal feedback it appears that the vast majority learnt at least a few new and potentially useful things and another session on Analytics is a possibility.
As soon as the training finished, I felt like death warmed up and started to understand why I hadn’t been feeling right over the past couple of days. It looks like the curse of my other half’s airport job has struck again. After a tricky drive home, I crashed out on the settee and was in bed by nine (unheard of for me). I reckon that was another factor in the training not going as well as I’d hoped because I must have been running on auto-pilot. Quite a few people on the team haven’t been well, so there’s definitely ‘something going round’.
Yesterday’s presentation was the first of a few I’ve got coming up over the next few months.
The annual OUG Scotland conference is in Glasgow next week and, as well as an opportunity to attend Jonathan Lewis’ masterclass on Monday and perform my first Oak Table ‘duties’, I’ll be presenting a new parallel query paper on Tuesday afternoon. The focus of this presentation is narrower but hopefully deeper than the previous presentations and I want it to perhaps be a little more practically useful. I’ll talk about how you switch on tracing, look at the individual trace files themselves more closely and explain some of the confusing results you might see. That last point is what drove me to write it in the first place and is based on some interesting email discussions with Jeff Moss when he was conducting his own experiments and feedback from previous attendees. Of course, Jeff can only make it on Monday 😉 but hopefully he’ll be able to see it at UKOUG in November.
Next up is the Miracle Definitely Very Last Ever Final Database Forum in Denmark. I’ll be giving a presentation in Tom Kyte’s Worst Practices stream (the idea being that if you know what not to do, it should help you decide what you should do instead). The current subject is Backup and Recovery worst practices but there are a number of related presentations on the agenda, so we might widen it to include other DBA subjects. I have no idea when I’ll be speaking yet, but it’ll be some time on Friday 29th or Saturday 30th September. I look forward to most of the conferences I attend, but I’ve never been to Denmark or to Miracle HQ (Mogens’ House?) so it should be good fun, if not liver-enhancing.
Then it’s a break to polish up both the OUG Scotland presentation and the How Many Slaves presentation for UKOUG 2006. I have to say that I’m particularly pleased that I’ll be able to get both presentations out of the way on the first day so that I can kick back and enjoy the rest of the conference 😉
OK, so this isn’t about your current presentations but… you should warn people to completely proofread any quotes they take from you.
We have a DBA who suggested PX to speed a slow query. He was looking for poor performers because the server IS ALREADY OVERTAXED! I went to quote your bullets of common sense and then I remembered.
Although you may be able to use Parallel Execution to make an inefficient SQL statement run many times faster, that would be incredibly stupid.
I don’t think that he’d get the fact that it was -you- calling him incredibly stupid and not me. 😉
… that would be incredibly stupid.
Mmmm, perhaps I should tone that one down a bit.
Maybe I could introduce …
Enabling PX when you server is already overtaxed would be incredibly stupid.
Although you may be able to use Parallel Execution to make an inefficient SQL statement run many times faster, that would be a waste of resources.
Does that sound better? 😉