A Cool Job

  • I saw recently that the newly approved TPC-E Benchmark, a test of OLTP systems, is having a test system submitted by Unisys. They'll be the first ones to send a system, in this case the ES7000, through it's paces with this benchmark.

    It's got to be a cool process and I've got feelers out to Unisys to try and dig further into what goes on when you try to build a system to test the benchmark. I have the feeling this will be an interesting story, so be on the lookout for that.

    However I'll speculate a little that this has to be a very interesting exercise. The application is static, after all, it's a standard set of routines (I'm assuming) that you'll run against your server. So the challenge is in sizing and tuning your system. Tweak a knob, hit the turbo switch, get advice from the MS guys, it's got to be an interesting day's work.

    And I'm sure it's incredibly enlightening. What does file placement mean? How far can you push a system? At the very least you'd think that they'd build some sort of real-time, F1-like dashboard so they can watch the system go.

    Heck, that would be a cool job just to build some application to receive the telemetry of the system in real time and display it on some sort of dashboard. Maybe I'll see if Unisys wants someone to work on that aspect of the test system.

    Now if we could just have some real-time knobs and sliders on the server to actually adjust various parameters...

  • Uh-huh... they need two special controls... one is a "Run SQL Faster" button and the other is a slide control to adjust the knowledge level of the developers writing code

    --Jeff Moden


    RBAR is pronounced "ree-bar" and is a "Modenism" for Row-By-Agonizing-Row.
    First step towards the paradigm shift of writing Set Based code:
    ________Stop thinking about what you want to do to a ROW... think, instead, of what you want to do to a COLUMN.
    "Change is inevitable... change for the better is not".

    Helpful Links:
    How to post code problems
    How to Post Performance Problems
    Create a Tally Function (fnTally)
    Intro to Tally Tables and Functions

  • LOLROTF.

     

    I need to put this up in my office .

  • In the 70's, my brother told me that the secret to buying audio equipment was in the knobs. The more, the better.

    Then the movie 'This is Spinal Tap' taught me that the really good ones are the ones that go up to 11.

    Unfortunately, as intriguing as the idea is, SOX has taught me that having the ability to tweak may violate certain SOD policies.

    And experience has taught me that screwing with the knobs isn't always the best way to achieve anything, although it has an amazing ability to achieve downtime. Gotta be related to Murphy who was, by the way, an optimist.

    So Steve, if someone invents what you want, who'll hire the troll to make sure that people keep their hands off the knobs?

    ------------
    Buy the ticket, take the ride. -- Hunter S. Thompson

  • We'll build a cage around it, like the missle launchers in "War Games" and give the key to the SOX guys

    You have a good point and I think this is part of why SQL Server has so many less knobs than Oracle and other systems. They're protecting us (or think they are).

    However during testing, or when you're trying to tune a system, it helps to have the knobs. I HATE hearing that something's changed and I have to trust the query optimizer rather than being able to experiment myself. We should be able to tweak this stuff in a QA environment and then move it to the prod servers.

    Maybe 2011....

  • I saw an interview with the real Murphy.  One of the things that had him in a knot was that he had not made a dime from the "law" while many authors and publishers had.  The interviewer pointed out that he had come up with the "law" while on active duty so the Air force would own the service mark (trade mark) anyway.

    ATBCharles Kincaid

  • Building the dashboard would be fun but it's still hard work. Designing the UI can be enjoyable but it would be the icing on a hard project. The work is in digging out the relevant data that truly means something. Distilling it down and presenting it in a fashion that efficiently communicates the situation. And providing the guidance on what should be changed. That's the kind of work that will keep you up at night.  

  • I think what Steve really wants is for that Prisus to have an F1 button on its dashboard.

    As in "Press F1 to try and pass something" 

     

  • Since they are not really production they could have used the Developer Edition instead of the Enterprise Edition of SQL 2005 with a cost saving of $374,912 ...

    RegardsRudy KomacsarSenior Database Administrator"Ave Caesar! - Morituri te salutamus."

  • The whole TPC exercise is a bit strange to me. It's almost like a bunch of jocks hanging around the locker room measuring each other's...well, you know. "Hey man, look at the size of his <insert phallic reference here>"

    But hey, there's no denying the bragging rights. "SQL Server - Fastest DBMS in the Universe"


    James Stover, McDBA

  • I think Bob wants a Prius

    Actually my Prius has plenty of buttons, knobs, etc.

    The TPC thing is contrived, no doubt about it. But I wish they'd take someone's app, something useful like a government records search app. Make it open source, let any government use it and then tweak the bejezus out of each set of hardware/software to make it scream.

    It's definitely like measuring something, and it's got to be a man that thought of it. I mean power tools, cars, body parts, who else wants to compare the biggest/fastest/most powerful else like men do.

    I think Unisys should just hire Tim Allen to be their spokesman. Can you see a 2 minute "Home Improvement" with the ES7000

  • There might be limit to "fastest".  I'm in mind of a humorous story written by Asimov about a chemical that had a ridiculously low solubility time (how fast something dissolves).  It was so fast that the chemical actually dissolved before the water hit it.  Then somebody figured a way to detect that the chemical had dissolved and prevented the water from hitting it thereby creating a temporal paradox.

     

    This was written during the time that Asimov was going for his doctorate in chemistry.  The review committee found out about it and they did not think that it was funny at all.  I’m such a geek that I laughed out loud while reading it.

    ATBCharles Kincaid

Viewing 12 posts - 1 through 11 (of 11 total)

You must be logged in to reply to this topic. Login to reply