IOUG-A Live! 2005 - Day Three, afternoon

This will be my final humorous blog entry from IOUG-A Live! 2005. I do intend to post a “final thoughts” post, but it will not be funny. Which won’t be any different than the rest of these entries, except by intent.

Prior to the talk on “Semi-joins” and “anti-joins”, I took one last stroll through the vendor room to check whether I’d won any of the prizes being given away. Surprisingly, I didn’t win (although I’m holding out slim hope for a couple of the drawings to be held after the conference). I did, however, get a chocolate latte thanks to my friends at the SAP booth, and strictly speaking, that meant more to me than an iPod. (To give you yet another example of how un-cool I am, I’m not yet 100% certain what an iPod is).

After the joins talk, it was off to Mike Ault’s discussion comparing Oracle on Solaris 9 vs RedHat 3.0 Linux. I didn’t get much out of it, but in fairness, I only attended it because there was nothing of particular interest or use to me during the hour, and Mike is generally an entertaining and informative speaker.

(By this point in the conference, I was really starting to wind down. Like a gourmand at a feast, I had reached the point where my brain was full. It was suggested I repartition my memory, or perhaps coalesce, but as I did not have a sufficient back-up strategy, I was unwilling to take the risk. I had done about 50 jumping jacks, hoping to compact the data, but to no avail).

Finally, it was time to head to my final educational session of the conf… er, symposium (I’ll get that right eventually), with the lyrically economical title “Advanced SQL Application Tuning: Find the Proverbial Needle in the Haystack”. As I walked into the room, I was struck by the fact that yet again, I’d accomplished something usually impossible: I’d once again managed to get my boss to pay me for a week while not requiring my presence in the office. If I could just find 51 more valid conferences, I could achieve the ultimate goal of any of the legion of Dilbert-esque workers out there, to get paid without that pesky requirement that we produce anything beyond carbon dioxide.

In the interest of truth (and because I know my boss has been known to peruse my blog), I have learned quite a bit this week, refreshed my memory on some things I’d forgotten, and refreshed my networking list of experts I can call upon when the fecal matter hits the high RPM rotating blades. It’s interesting to me that “networking” and “going out drinkin’ with the guys” are indistinguishable to the casual observer, the only measure being how quickly someone with more intelligence than you can be contacted when all hell breaks loose. Nevertheless, I’d successfully networked to the point of inebriation with several of the biggest names in the Oracle DBA and development world, and at the end of the day, the only thing better than that is that it was all covered by my conference admission fee.

Anyway, back to the final session of the conference. The long and the short of it is that I’m as important as I always felt I was. One might think this is an odd and rather self centered conclusion to take from a session that never mentioned me once, but I’ll explain. The speaker contends that there are five levels of tuning, and while most sessions (this is true in my experience) seem to focus on the first three (RDBMS, Hardware and Operating System), that in fact the best “bang for the buck” is usually found in tuning inefficient programs and poorly designed queries, and that’s where I come in. If I were not out there, writing inefficient programs and poorly designed queries, there’d be nothing for you DBAs to tune, no miracle worker reputation for you to gain. Don’t thank me, I just do it because you’re that important to me.

Seriously, as a developer (and database designer) who understands the importance of proper query tuning as well as coding efficient algorithms(*), it’s always an ego boost to sit through an hour of someone telling me how important the core focus of my entire career is. (I should say, this was one more session presented by a DBA trying to blame developers as the source of all bad SQL, and while that may be true, I still contend that a developer who produces bad SQL is a bad developer. Just like a DBA who can’t keep his… database up is a bad DBA. There are inept people in all walks of life, stop assuming that I am one just based on my job description. Please, it's much more valid to assume I am one because of the incompetent blather on this blog.)

(* Named after Al Gore, obviously)

And with that session, all there was left to do was attend the closing session (“10g, Do You Really Need A DBA Anymore?”). As this was a panel consisting primarily of DBAs, the surprising conclusion was that yes, you really do, quite the opposite of what we developers know to be true. (If I’ve managed to fan the flames, just a little bit, of the DBA – Developer war, I will have done my job).

And actually, I don’t really believe that little dig. DBAs are very important, for any number of real reasons, but I think the most important one is so that we developers have someone to blame poor performance on. DBA, contrary to popular belief, is an abbreviation for the Latin words for “scapegoat”. It’s true. Look it up. By the time you’re done, I’ll be gone.

And then the spontaneous coalescing of nerd power began to disperse, spreading out across the country and across the globe, having learned (if nothing else) a few more key buzzwords with which to bamboozle their management into further raises. I will miss them, these Techies in the Mist.

Copyright © May 4, 2005 by Liam Johnson. http://www.liamjohnson.net