Tuesday, October 25, 2016

SQL Pass Summit 2016... day 2

Day two in Seattle started early....very early....after turning and tossing since 2 AM, I decided to get out of bed at 3 AM. I read a little, hit the gym, showered and wrote the following blogpost: SQL Pass Summit 2016... day 1.... just looking around
At 7 AM it was still pitch dark, Here is a pic of what it looked like at that time

I went down to the lobby, met a co-worked and we walked to the Washington Convention Center. We had breakfast, we were joined by another co-worker, we also talked with some people at the table.

After breakfast, it was time to go to our sessions. I went to the Locking, Blocking, Versions: Concurrency for Maximum Performance [DBA-331-PC] session by Kalen Delaney

This is the description of this session

Failing to design an application with concurrency in mind, and failure to test an application with the maximum number of expected simultaneous users is one of the main causes of poor application performance.

SQL Server offers two methods for an application to provide data consistency: Pessimistic and Optimistic Concurrency Control. In this seminar we’ll discuss what consistency might mean to your applications, and describe the details of how each of the concurrency models works internally. We’ll examine the costs and trade-offs between the two concurrency models that SQL Server 2014 supports through different transaction isolation levels. Finally, we’ll examine the tools available for analyzing and troubleshooting blocking problems, including metadata views and Extended Events. You will learn:

• why concurrency management is a crucial part of database and application tuning
• how each of the transaction isolation levels impacts concurrency
• the details of how SQL Server implements pessimistic concurrency through locking, and when locking causes blocking
• how the snapshot-based isolation levels implement optimistic concurrency, and the problems to be aware of
• how In-Memory OLTP allows maximum concurrency with no locking
• how to use the metadata views and Extended Events to troubleshoot concurrency problems and monitor the resources used to manage concurrency.

Here is a picture of Kalen presenting

The session was pretty good, there was some stuff I knew, some stuff I didn't know but should have known, some stuff I knew at one point but forgot. We had an one hour break for lunch and I must tell you the lunch was excellent, here is also a pic of what I ate. You can see, there is some salad, carrots, potatoes, stuffed tomatoes, salmon, pork and also some rice

After lunch we went back to the session. I was there a little early and decided to play around with some code. I came up witht the following teaser, I then posted it on this blog SQL Pass Summit... SQL Teaser Transactions

The session ended at 4:30 PM, we then walked back to our hotel to take a one hour break. We met up at 6:15 PM in the lobby and went back to the Washington Convention Center for the PASS Summit 2016 Welcome Reception At the welcome reception I saw many old friends like Brent Ozar, Mladen Prajdic, Jason Strate, Thomas LaRock, Rob Farley, Tim Ford, Kevin Kline and many more. Sometimes I miss being a SQL Server MVP, because these are the folks I hung out with back in the day.  There were a ton of people at the reception, here is a small selection of them.

While being at the PASS Summit 2016 Welcome Reception I left my iPad in my hotel room to record a time lapse, this ran between 6:15 and 8:30, here is what that looks like


No comments: