Q&A with the SSIS Design Patterns Author Team
SSIS Design Patterns
Q: If your pre-con had a theme song, what would it be and why?
Andy Leonard: For my theme song, it would have to be something by Toby Keith – probably "As Good as I Once Was."
Jessica Moss: “Let’s Work Together” by Canned Heat. The pre-con speakers have been tirelessly working together to put together a full day of the best design patterns we can muster up. Attendees will get to learn design patterns from every one of the presenters throughout the day.
Matt Masson: “The Final Countdown” by Europe, because it is overly dramatic, showy, and reminds me of “Arrested Development” every time I hear it.
Michelle Ufford: My pre-con theme song would probably be “Fuel” by Metallica, because we're going to discuss strategies for speeding up the movement of large data.
Tim Mitchell: Without a doubt, it would be the "I Like to Move It" song. (You know, from the Madagascar soundtrack. You're singing it now, aren't you? You're welcome.) Moving large volumes of data around in a manner that makes it easy to troubleshoot, audit, and consume the results takes skill, but doing it with style takes passion. We all like to move it.
Q: Andy, what excites you most about today’s SQL Server Integration Services?
A: I like the direction represented by SQL Server 2012 Integration Services. The SSIS Team at Microsoft responded to suggestions from the data integration community and made a good product even better. Examples include upgrading packages from SSIS 2008 R2, backwards compatibility, and the SSIS Catalog.
Q: Jessica, what's the most surprising statement attendees might hear you say during the pre-con?
A: Besides SQL Server, I’ve worked with DB2, Oracle, mySQL, Access - and SSIS can connect to all of them!
Q: Matt, what’s the biggest myth surrounding SSIS that you’d like to debunk?
A: The biggest myth I love debunking is that because SSIS is a fraction of the cost of other ETL platforms, it can’t provide the same functionality.
Q: Michelle, what still trips you up in building SSIS solutions?
A: Robustness. There's a lot that can go wrong in an ETL environment. Data types can change upstream, unexpected values can cause conversion issues, the destination server can go offline… That's why it's critical that ETL packages fail gracefully and provide visibility when something unexpected happens. We'll be discussing both of these topics during our pre-con session.
Q: Tim, if attendees could start putting into practice just one thing after this pre-con, what would you want that to be?
A: Be a pessimist. There's plenty of time for optimism when participating in the sales cycle, when awaiting your yearly bonus, or on #sqlkilt day when you hope Andy Leonard doesn't arrive showing leg. When designing ETL, assume that all inputs and outputs are volatile. Assume that all data is bad. Know that every process you build will eventually encounter a speed bump or even an all-out train wreck. If you design your ETL processes in such a way that data anomalies and errors are consumed and handled as part of the process, you'll have higher quality data that is more consistently available, and you'll get fewer middle-of-the-night wakeup calls.
Check out our other Q&As with PASS Summit 2012 pre-con speakers.