SQLServerCentral Article

The Mile-High View From PASS 2007



Those lucky enough to attend this year’s PASS 2007 Preconference were treated to two full-day seminars of high-quality content from a host of Microsoft MVPs and other high-caliber professionals.  The presenters included prolific SQL Server authors Itzik Ben-Gan and Kalen Delaney, as well as Microsoft MVP Andrew Kelly.  The preconference provided attendees with the kind of valuable insights we’ve come to expect from these experts we’ve come to depend on in print and in the newsgroups.

The only downside was that the preconference seminars were full-day events held over the course of two days, making it very difficult to choose which seminars to attend.  On Day 1 I chose to split my day in two and attend two seminars.  I started my day with the “Introduction to Microsoft Business Intelligence Seminar”, led by Curtis Young.  Young is a Senior Consultant for Performance Management COE (Center of Excellence) for Microsoft Consulting Services (MCS).  His discussion was focused primarily on the DBA side and focused largely on the Big Picture view of Business Intelligence (BI) in the enterprise.  Young began with a very high-level overview of Business Intelligence (BI) and provided valuable insights into the reasoning behind the design of Microsoft’s BI products.  He also covered best practices in BI solution design, with a solid discussion of the problems that result in a lack of user acceptance for BI solutions.  Though not overly technical, Young’s seminar provided a very interesting and informative discussion of BI solution design.

After lunch on Day 1, I slipped into Itzik Ben-Gan’s standing-room only seminar called “Inside T-SQL Querying, Programming and Tuning – Putting Your Knowledge Into Action”.  Originally planned for approximately 125 attendees, Ben-Gan’s developer-oriented discussion was packed with over 200 attendees.  The discussion covered a wide range of development topics, including such diverse areas as optimizing T-SQL queries on SQL Server 2005, SQLCLR programming best practices, and SQL Server 2008 new features.  The attendees sat riveted in their seats as Ben-Gan provided valuable insights into new SQL Server 2008 features like the MERGE statement, grouping sets, row constructors, and the HIERARCHYID data type.

On Day 2 I attended Kalen Delaney’s seminar entitled “Controlling and Reusing Query Plans”.  Delaney’s seemingly effortless style of teaching made this seminar of a somewhat complex topic very easy to digest.  She covered topics including analyzing query plans, plan caching and reuse, overriding the optimizer, and plan guides.  As Delaney explained, there are many different variables that affect your query plan efficiency, caching and reuse.  Managing those variables is the key to making your queries as efficient as possible.  Overall, this was an excellent presentation by a true SQL Server guru.

The preconference ended with the PASS Welcome Reception featuring the Special Interest Group (SIG) Bowl later in the evening.  The SIG Bowl was a gameshow-style meet-and-greet/quiz event in which audience members and contestants won a wide array of prizes.

I have to say that overall I was extremely impressed by the quality of content provided by the preconference presenters as well as the expertise and professionalism throughout the preconference.  In the next article I’ll cover the actual PASS Community Summit, which began on Wednesday.

Day 1

The 2007 PASS Community Summit in Denver officially began on Wednesday morning (September 19), with the keynote address being delivered by Ted Kummert, Microsoft's Corporate VP of Data and Storage Platforms Division.  The keynote began with an overview of SQL Server nostalgia for us old-timers, but quickly moved into Kummert's impressions of where SQL Server currently stands and where it is headed.  Here are a few takeaway facts about where SQL Server stands now:

  • SQL Server is currently the unit leader in the database market [IDC]
  • Microsoft is currently the fastest growing Business Intelligence (BI) vendor [IDC]
  • There are 619,000 SQL Server-trained IT professionals world-wide
  • SQL Server has set new records for scale and volume on the TPC/H and TPC/E 1 TB and 3 TB benchmarks

And for the future, Kummert said that Microsoft will incorporate 172 customer requests into SQL Server 2008.  During the keynote Kummert invited some of Microsoft's top talent to demonstrate key features of 2008.

Shishir Mehrota, Director of Program Management for SQL Server, gave demos on the elusive SQL Server 2008 spatial data types and on a new feature at the top of every DBA's wish-list: the Declarative Management Framework (DMF) and Resource Governor.  The DMF allows DBA's to use declarative methods to set system policies.  How many DBA's have wished that they could put a stop to developers creating objects in the default dbo schema?  With DMF that power is just a few mouse-clicks away.  The Resource Governor allows the DBA to dynamically assign SQL Server workloads to resource pools and restrict resource-intensive operations from holding a server's resources hostage.  Mehrota even managed to slip in a quick demo of the new SSMS Intellisense feature, a potential boost for developers.

Microsoft's Senior Product Manager for Windows Storage, Jason Buffington, demonstrated the Microsoft System Center Data Protection Manager (DPM), which promises to provide continuous recovery for SQL Server and just about all other Microsoft server products.  DPM will allow DBA's to set recovery policies for both tape and disk that will result in no more than 15 minutes worth of lost data from any given point in time.  That's about as real-time as it gets folks!

Of course what SQL Server 2008 demo would be complete without the obligatory LINQ demo, brought to PASS by Erick Thomson, Microsoft Program Manager.  While LINQ is an exciting technology, I for one want to see something a little more exciting than a standard LINQ query and an automated simple join.  Microsoft definitely needs to up the ante a bit to demonstrate the true power of LINQ.

I was unable to get the last presenter's name and title (I apologize), but his demo of the Entity Framework Designer, the long-overdue Data Profiler, and a new dimension editor, should excite the BI community.  All of these represent standard tools of the trade in the BI world, and I can't wait to get my hands on them.

Overall, the keynote on Day 1 did what any good keynote should do:  It helped build excitement for the event, the next release of SQL Server, and demonstrated the next generation of tools of the trade for SQL Server professionals.

Day 1 Sessions

The 2007 Session Guide was absolutely packed with interesting and intriguing technical sessions.  I couldn't wait to dive in and get some solid technical information from the experts.  Normally I would have focused my energy on database and application development track, but decided to rotate my schedule around the various tracks to get a more complete picture of what was going on with my favorite database product.

To start with I jumped into the "Microsoft Excel 2007 BI Capabilities" session on the BI track.  Presented by Rushabh Mehta and Alejandro Leguizamo of Solid Quality Mentors, this session provided a very informative look at the new BI features of Excel 2007.  Let's face it – no matter how well you design your databases, user interfaces, cubes, and reports, business users will still end up using Excel.  It's what they know and what they're comfortable with.  Microsoft has decided to capitalize on that knowedge and pervasiveness in the business world by making it easy to connect Excel to SQL Server Analysis Services.  Now your users can literally drag and drop dimensions and facts into the Excel pivot tables they're already familiar with to generate customized reports, complete with drill-downs.

The next session I attended was the "Deep Dive Into XQuery and XML in Microsoft SQL Server", a very good session led by the ultimate authority on SQL Server XML, Microsoft's own Project Manager for XML in SQL Server, Michael Rys.  Rys' session began with an enlightening overview of when to use XML, when to use relational storage, and when to use XML plus relational.  He quickly dived into the optimization of XQuery in SQL Server and gave a detailed examination of best practices.  Some of the information he gave was surprising, some (particularly in the best practices portion) was completely unexpected.  While the XQuery engine actually manipulates your queries to leverage the power of the SQL Server relational engine, how you write your XQuery queries can have significant performance implications.  Rys' information will prove invaluable to developers and DBA's trying to squeeze performance from the XML data type.

My next, and final, session of the day was "Creating an SSIS Auditing & Execution Framework", by author and MVP Erik Veerman of Solid Quality Learning.  Veerman began by demonstrating that even as the use of ETL solutions has skyrocketed over the past few years, the Total Cost of Ownership (TCO) of those solutions has also increased dramatically.  This laid the foundation for his proposed solution—creation of a common ETL framework.  Though it is a very simple idea that's been adopted in most areas of development, there are still thousands of brand-new one-off "temporary" solutions popping up every day.  And as all developers and DBA's know, those "temporary" often become a permanent fixture.  Veerman demonstrated the "Project Real" solution, which provides templates and examples for auditing and execution of SSIS packages.  (Project Real is available for download at http://www.microsoft.com/sql/solutions/bi/projectreal.mspx.)  He also demonstrated his recommended method of using database snapshots to quickly roll back an ETL process gone awry.

Overall, the first day of the summit turned out to be a great learning experience and a terrific time for many SQL Server professionals.  The quality of content exceeded my expectations, and I can't wait for tomorrow's sessions.


5 (1)




5 (1)