SQL Saturday

SQL Saturday has been running for a number of years around the world, I went to the first one ever held in Australia. It had a huge turnout with around 20 different events to choose from over the day. It had a fair few sponsors present as well as lots of Microsoft Valued Professionals (MVPs).

sqlsat296_web

Of the 20-ish sessions to choose from I went to these:

 

1st session: Kick-ass Ad-hoc Cube Browser build in SSRS 2014

I’ll freely admit I went pretty much for the session title alone. Creating MDX driven reports from SSRS has never as smooth as should have been so I did have some genuine technical interest though.

Grant Paisley showed his Reporting Services generic cube browser report. This was a nice way to have a single report that offered enough parameters you could create virtually unlimited reports from it without having to create new reports. A huge time saving win if you are working with SSRS and Analysis Services.

He has a site that hosts lots of free SSRS report templates you can can try online:  http://www.reportsurfer.com/ReportLibrary.htm

 

2nd Session: I have the data now what! (Troubleshooting problems using logs)

The most mundane sounding session but pretty interesting so bear with me on this one. The demo by Henry Rooney walked us through using LogParser http://technet.microsoft.com/en-au/scriptcenter/dd919274.aspx. This a free Microsoft tool that basically puts a SQL queryable interface on lots of different Windows & SQL event logs. Being able to use SQL to search for the text “Failure” across Application, Security, SQL Server & Cluster logs is a lot better than manually sifting through the Windows event logs for strange looking events around the time something broke. Even better – you can automate this and get it in a daily report.

 

Lunch was spent with 11 MVPs discussing all things SQL Server while eating the largest apple I’d even seen.  Questions were asked about Parallel Data Warehouse (PDW) as Matthew Winter was there (My notes on his presentation at Microsoft a few months ago:https://uxcpsim.wordpress.com/2013/12/19/parallel-data-warehouse-sql-servers-big-brother-presentation/ ).  Most discussion was a bit DBA-ish for me but it did strike a chord when there was mention of the difficulties of trying to track down performance issues on a shared SAN. I think a lot of people in the room may have had a little cry at some stage over that.

 

3rd Session: The Incredible Shrinking Execution Plan

Back in developer territory, Rob Farley suffered a gargantuan bout of demo-laptop-fail but managed to get back on course to show the evils of using table valued functions. Summary: these execute per row so are slow where-as in-line table value functions are set based and can be fully utilized by the query optimizer. See a related article here: http://blog.waynesheffield.com/wayne/archive/2012/02/comparing-inline-and-multistatement-table-valued-functions/

Mental note: if you are writing a view that has a begin and end statement you probably need to reconsider what you are doing.

 

4th session: Event notifications for the proactive DBA

Seeing as how I normally work in development databases with no DBA coverage to speak of, having another free tool to help figure out why the application has just stopped working is always good. Martin Catherall  walked us through using the built in SQL Server Event Notification mechanisms with Service Broker to enable asynchronous handling of events that are placed on a queue. Getting told when deadlocks happen and disk space runs out is valuable not just on production environments.

 

There was a lot of information to digest in just one day but a great opportunity to stay ahead of the game technically and do some quality networking. If you can’t make the regular user groups in the city during the week then this is a great way to get all that content all in one hit.