Author: way0utwest

  • T-SQL Tuesday – I/O Fallacies

    It’s time for the March T-SQL Tuesday and this month Mike Walsh picked the topic. It’s I/O and you can read about his idea on the Straight Path Solutions blog.

    I am not really a hardware guy. I like it, and it can be interesting, and in the past I spent a lot of time learning about SCSI, PCI v EISA, RAID, etc. and it’s impact on SQL Server. However about the time that SANs became popular I found that I couldn’t keep up with all the technologies that were coming out and it wasn’t a part of my job. The companies for which I worked typically had someone assigned to deal with hardware, and often just storage.

    So I let them figure out what hardware we needed, and what implementation. My only requirements were that I have xxx sized drives (or arrays/LUNs) for my SQL Server to separate out data from logs or backups. I would even let them know that we wanted to support yyy reads/sec of zzz size, or similar specs so that they could design the subsystem.

    One day we got a new SAN.

    I needed new storage for a SQL Server and soon the storage guy came to talk to me. He said that the vendor basically set up the entire SAN as one large array of disks and he would just carve out xx GBs for each of my drives. I wasn’t pleased with that and requested a meeting with the vendor. I got 2 sales guys and a sales engineer to talk to me, none of whom had ever really spent time using the product and were repeating what their engineers had (supposedly) said.

    Needless to say I disagreed with them, but I let them do things their way and we ended up with performance problems at times on the SQL Servers connected to the SAN. Not all of them, but a few that were very busy at times had delays getting I/O satisfied if other servers also needed lots of I/O.

    I left shortly thereafter, so I’m not sure what happened with those servers. However I think the lesson here is that a SAN doesn’t magically create better storage. A SAN can often handle higher I/O loads overall than DASD (Direct Attached Storage Devices), but it’s a balancing act for the SAN. Like an Ethernet network, often the SAN is counting on a distribution of I/O loads. Not all servers needing lots of I/O at the same time.

    Or if they are, that the write cache can handle things, or the servers will retry their requests (or wait). For file server type operations this can work great. But for SQL Server it can be an issue. And write caching transactions that SQL Server expects to be on disk and permanent, or Durable in ACID terms, can lead to corruption.

    Be careful of thinking a SAN can remove all I/O worries from your storage design. Good fundamental disk practices are still required.

  • It’s Like the MVP Summit

    I missed the 2010 MVP Summit, which is disappointing. There were a lot of SQL 11 talks and I was looking forward to those. I was also hoping to get together with a lot of MVP friends that I rarely see.
    However this weekend in Charlotte, for SQL Saturday #33, it appears that it’s like a mini-summit, and thanks to Peter Shire and SQL Sentry for assembling an amazing lineup. I am sure I am forgetting someone, but look at this lineup:

    • Andy Warren
    • Andy Leonard
    • Jessica Moss
    • Kevin Kline
    • Rushabh Mehta
    • Wayne Snyder
    • Allen White (the roommate I missed at the Summit)
    • Andrew Kelley
    • Aaron Bertrand
    • Tim Ford
    • Denny Cherry
    • Geoff Hiten
    • Ashton Hobbs
    • Alejandro Mesa
    • Mike Walsh
    • Patrick LeBlanc
    • Kendal van Dyke
    • and more

    Not all MVPs, but a few probably deserve it.
    I think I’m the weakest link of this bunch. This was one of the best lineups I’ve seen at an event, and incredible for a free event. SQL Pros in Charlotte, you should consider yourselves lucky to have the chance to attend this amazing day of training for free.

  • Opening Keynote – SQL Saturday #33

    I didn’t quite read this, but this was the written part of what I have as the opening keynote for SQL Saturday #33.

    Welcome to SQL Saturday #33 here in Charlotte.

    My business partner, Andy Warren, has had a ton of ideas over the years. He likes to remind me that they might not all be good ideas, but there’s no shortage of them. A few of them stand out, like his idea for the Best of SQLServerCentral books. I was driving home in a rare Denver rainstorm one night when he told me we should republish the best articles from the year in a book format. I had to pull over and debate it since I was sure no one would want a copy, but I was proved wrong. That series has been very popular over the years and we are on our eighth volume this year. SQL Saturday was another one of idea I didn’t believe in at first

    When we started our training business, SQL Share, we were looking for ways to market the business. We tried some of the traditional methods of advertising, and had even considered purchasing a booth at the PASS Summit or TechEd. Those were expensive, and we were not sure they would be good investments. In our market research, which is really just a fancy term for us talking to regular people like all of you, we realized that the vast majority of IT professionals never attend conferences and other large events. Lots of people rarely go, usually once or twice in their careers, and a very lucky, select few go quite often.

    When Andy approached me with the idea of running free events to get more training out to those that can’t go to conferences, I was skeptical. I wasn’t sure that it would even run, much less help us market our business, but Andy had proven me wrong before and both Brian Knight and I lent our support.

    And money.

    Over many phone calls, we worked on the idea and eventually SQL Saturday #1 in Orlando was held three years ago. Since then dozens of events have been held all over the country, and while they haven’t necessarily resulted in a lot of monetary profit, they have brought us a lot of satisfaction, and I’m proud to be associated with them. I’ve helped in a small way, but most of the credit has to go to Andy Warren.

  • Heading to Charlotte

    I think I’ll forever associate Charlotte, NC with my middle son. We were heading back to Denver from Virginia Beach a few years ago after a vacation. Our flight was scheduled to go through Charlotte before heading to Denver. My daughter and my middle son were sitting with me as the plane sped down the runway in Virginia Beach when about halfway down the pilot slammed on the brakes. If you’ve never experienced that, it’s very unnerving. My daughter gripped my arm as my son looked up and said

    “Are we in Charlotte?”

    Totally serious since he’d been engrossed in a book. My daughter, who looked like she might cry up to this point started laughing. She told him we hadn’t even taken off.

    There was a problem with the plane, and we ended up spending another night with grandma, which pleased her and the kids. Everything was fine, but it is one of those events that sticks with you.

    This morning I’m heading to Charlotte, no kids, to SQL Saturday #33, to speak and see some friends. I’ll be giving my Modern Resume presentation and also part of the keynote presentation. Andy Warren and I, along with PASS President Rushabh Mehta, will be up to talk a little about the history of SQL Saturday and it’s current transition to PASS.

    It should be a great event, and the list of speakers is impressive. Kevin Kline and Joe Webb from TN, Denny Cherry from CA, Jessica Moss and Andy Leonard from VA, Aaron Bertrand from RI, Patrick LeBlanc from LA make this one of the more impressive events outside of the national conferences.