Monday, November 19, 2012

SQL in Seattle

This past week, I went to Seattle to attend the SQL PASS Summit. I had a great week learning about SQL Server, exploring Seattle, and networking with my fellow SQL Server professionals.

Pikes Place Market
You can't talk about Seattle without mentioning the weather. Personally, I found the weather to be fairly nice. Recently, I had purchased a new rain jacket for the Austin City Limits Festival, and was looking forward to putting it to use in Seattle. Funny, I left Texas only to have it rain while I was away. In Seattle, it only rained once, any only for a few hours. Not really that cold either, a fleece was all you needed.


SQL Server 2012 Service Pack 1 has been released. It's time to update your servers. If you've been sitting on the sidelines, no more excuses.

Heckaton, as in you have a heck-o-ton of data and need things to be quick. This is an in-memory database technology that promises to greatly improve performance.

Column Store Indexes will now be updateable and able to be the clustered index. This will help performance in Data Warehouses. Nice.

And perhaps the biggest announcement...

Business Analytics PASS, aka BI PASS

There will be a BI-focused Data Analytics Conference in Chicago next year. I think this is a good thing. I like BI, but it is has grown to the point where it needs its own conference. However, I would like to see PASS retain 10-20% of BI sessions at the Summit. At the same time, BI PASS should also retain a similar percentage of DBA content.

SQL Family

One of the best parts of PASS is reconnecting with old friends and making new ones. I especially enjoyed seeing my old SQLskills classmates and SQL Cruise shipmates. Because of the intimate nature of these two courses, you build a solid relationship with your peers. This was very evident seeing each other at PASS over a year later and picking up right where we left off. It seemed everywhere I turned, I was running into someone from one or both of these two classes.

I met so many new people from countries all around the world. One interesting person I got to spend some time with was Benjamin Nevarez. If you haven't checked out his book, Inside the Query Optimizer, go grab a copy and prepare for your brain to melt.

Session Highlights

There were so many cool sessions, but these were some of my favorites...

Klaus Aschenbrenner had a pre-con where he took you through his Trouble Shooting Methodology. I enjoyed how he showed us some of his performance tuning techniques and a few different ways to deal with SQL problems.

Got Security?
Andreas Wolter is a security guru and he did not disappoint. In this breakneck session we took us through security basics, a few encryption gotchas, and got into some advanced topics. This was a lot to squeeze into a seventy-five minute session, so I hope to see him in a half-day session in the future.

Both Andreas and Klaus were in my SQLskills rotation last Summer. So when I saw their names on the schedule, they were some of the first sessions I picked. Both are top-notch SQL guys. If you get the opportunity to see one of them, go for it!

Adam Machanic took us deep into parallel execution plans and showed us how to tell what was going on and why. Even if you don't have parallelism issues, everything we learned can be applied to 'normal' query plans and tuning, as well. This was my favorite session of the week.

Waddayamean the server failed over?

Brent Ozar had a couple of nice sessions this year. In one, he discussed some of the projects where he has used SQL Server 2012 Availability Groups. He spoke about the need to test your DR plans befre disaster strikes so you will know if they work. We were able to discuss some real world examples involving the Stack Exchange sites and Hurricane Sandy.

Renaissance Festival Anyone?

Later in the week, Brent demoed a new version of sp_blitz which adds Plan Cache Analysis to the tool. I enjoyed that he actually went behind the scenes and explained what the code was doing instead of just showing us the finished product.

Jason Strate had an excellent session where he showed us how to dig into the Plan Cache using XML Queries. I love using the Plan Cache, and feel it is an often overlooked area of performance tuning. He took it to the next level showing us how to do this more efficiently using XML Queries. I've been chatting with Jason for a while over Yammer, so it was great to finally meet him face to face.

Suffering from Success

I'm starting to think PASS Summit is suffering from its own success. This year is the largest PASS Summit yet, but I think it has gotten a bit unmanageable. The attendance is large enough that the session rooms are spread out on opposite ends of the convention center.

Many of the sessions I tried to attend were filled to capacity and I was not able to get in. In one case, I had to go to my 4th choice session. For some of the popular sessions, people were lining up 30-45 minutes early. So if you wanted to guarantee you had a chance to see some of these sessions, then you either had to leave the previous session early, or skip it altogether.

To make matter worse, several times there were popular, filled sessions in smaller rooms; while the large room 6E was empty. I would understand the occasional miss, but this seemed to happen quite a few times.

Furthermore, with so many people it was difficult to navigate the crowds or find time to meet new people. Everyone seemed to be rushing to their next session so they could get a seat.

When I consider all of this, I think having a dedicated BI-focused conference will help alleviate the crowding problems that I experienced this year.

See You Next Year in Charlotte

Gimme the IOPS and no one gets hurt!
Next year, Pass will be held in Charlotte, North Carolina October 15-18. I think it will be cool to check out a new city for PASS, but I am a little bummed about the dates. It is coming right on the heels of the Austin City Limits Festival so it might be difficult for me to attend. We shall see...

No comments:

Post a Comment