PASS Summit 2009 Conference – Day 1 Recap
Posted by denglishbi on November 4, 2009
So the first day of the PASS Summit 2009 Unite conference has come to an end. Overall we are off to a good start and the opening remarks by Wayne Snyder were full of interesting statistics. This year 42% of the 2200 attendees are first timers and 31% of the attendees are here for data warehousing and business intelligence (28% DBA and 20% App Dev). There are 400 Microsoft representatives present at the conference and 98 SQL Server MVPs. Very impressive numbers and based on last years attendance of 2445, I was very surprised, considering that in 2007 there were only 1528 attendees.
Some of the other items that were talked about were FastTrack 2.0 being announced since the initial release of this back in February this year. The material has been updated and IBM has now been included with three new reference configurations. Report Builder 3.0 will support integration with scale out data warehousing as a source and also include Report Gallery capabilities (reuse items from a repository like charts), sparklines, and data bars. Some new benchmarks were set with SQL Server 2008 R2 (OLTP 2,012 tpsE TPC-E and DW 102,778 QphH TPC-H 3TB) which should be available possibly next spring (or at least 1st half of 2010), CTP release of SQL Azure (RTM beginning of next year), and SharePoint 2010 lists and libraries will now be made available as services that can easily be consumed by tools like PowerPivot (formerly ‘Gemini’).
As far as the sessions that I checked out, here are some items that were discussed that might be of interest for each of these sessions:
- BI with Excel 2010 & PowerPivot for Excel: in-memory column compressions (distinct values), data stored in Excel (so you won’t want to be emailing these files around to users –> leverage SharePoint 2010 for delivery and consumption), PowerPivot = massive cache of data within tables, DAX (Data Analysis Expressions) will be new expression language available to leverage, ability to search within filter selection now and filtering capabilities are available as embedded filters back in the PivotCharts again, PowerPoint Gallery provides ability to preview your files before you open them and you can manage/configure the refresh rates
- What’s new in PerformancePoint Services 2010: items dropped from PPS 2007 were Planning, OWC (use Excel Services now), ODBC data source (use BDC –> now BCS instead), trend analysis and the items added were SharePoint Service Application, Decomposition Tree, enhanced scorecards, browser support (firefox and safari), increased MUI support, accessibility, analytical pie charts, KPI details report view, filters are now separate content types (can reuse them across dashboards). Other items include support for SSAS conditional formatting, can expand/collapse hierarchies in scorecards now, autoselect members in scorecards now (support for formulas –> no longer have to be static members), Preview site is gone, increased filtering and sorting support with scorecards (just like Excel), preview button in analytical reports is gone (access through right-click menu), silverlight utilized for Decomposition Trees and these are not a report type, just available through analytic items and scorecards (data bar display in the tree for additional visualization), browsing dimensions in details for items in Dashboard Designer now supports hierarchies instead of just providing a long field list like Excel 2003, ability to migrate PPS 2007 content from PPSMonitoring repository, data sources can be setup for Per User, unattended, or unattended/authenticated user (custom data) within each data source now, other application settings can be accessed through SharePoint 2010 Central Admin within Application Settings.
- SQLBI Methodology: Marco Russo and Alberto Ferrari were a riot and I loved their skit presentation. Great presentation and very informative. Leverage a combination of Inmon and Kimball using relational data warehouse with natural keys and data marts with surrogate keys which will then be leveraged for SSAS cubes (OLAP). All logic and rules implemented in ETL and simply consume views as additional layer for reporting and DSV reference.
- Automating your Data Warehouse: some interesting ideas and liked the demo of setting up AMO logic to be able to pull in the OLAP QueryLog information to be able to automate the Usage Based Optimization to generate aggregation designs for partitions. Also use this to be able to programmatically validate SSAS database without using BIDS. Can also use this to update counts on objects within the SSAS database and to archive and remove old partitions that are no longer needed.
I also checked out a HOL of Master Data Services to take that product for a ride. UI navigation seems to be a slight issue, so just trying to get used to the new product offering. I was able to create a model, setup entities, attributes, hierarchies, import and consume data.
The Microsoft BI stack leveraging SQL Server 2008 R2, SharePoint Server 2010, and Office 2010.