Today was the 1st day of the SQL PASS Summit and the keynote was full of some great announcements and demos. The keynotes are a massive with almost all of the 3894 conference attendees from 57 different countries.
All of the keynotes from this years conference not only are the keynotes being streamed live, but many of the sessions are also being streamed live via the SQL PASS website on PASS TV. The full schedule can found on the PASS TV page.
Bill Grazino told us turning his keynote that pass has sponsored 9 non-PASS events around the world. Including these 9 events, pass has provided well over 500,000 hours of training to members of the community, much of it like SQL Saturday’s, User Groups, etc. being available to the attendees free of charge. PASS has created a business analytics specific conference which will be held April 10-12 2013 in Chicago, IL. More information will be available on the SQL PASS website shortly.
When Ted Kummert took the stage he talked about some of the great products that Microsoft has released this year including SQL Server 2012, the Surface, Windows Server 2012, Office 2013 as well as many others.
Ted announced that SQL Server 2012 has been released as of today. This release aligns with Office 2013 and the new SharePoint 2012. Ted also announced project Hekaton which is a new in memory database engine which will be introduced within the next major release of SQL Server. Project Hekaton is a fully in-memory database engine which will greatly improve OLTP performance. While this new engine is a part of the SQL Server database engine, this portion has been totally rewritten to really maximize the hardware under the SQL Server database engine.
Including Product Hekaton is the Hekaton AMR Tool which (when renamed) will help identify bottlenecks which can be resolved by converting the database to a Hekaton database engine. Because the Hekaton engine is running with the entire database being in memory latching has effectively been removed. Hekaton also supports recompiling stored procedures from running T-SQL to running as native machine code. This allows the stored procedures to run much, much faster without any actual code change. While not all workloads are going to be right for the Hekaton database engine, but those that are should see a massive performance improvement. Because project Hekaton runs within the same database engine that we have today if you know SQL Server you will already know how to work with a project Hekaton database.
Another great announcement is the clustered column store which will be introduced in the next major release of SQL Server. Along with the clustered column store index we will also be able to update tables which have column store indexes created on them, no matter if the column store index is a clustered column store index or a non-clustered column store index. This will introduce massive performance improvements to real time data warehouses and make data warehouses much easier to setup and performance tune; even with the largest workload.
For the non-relational side of things you can today download the Microsoft HDInsight Server CTP which is Microsoft’s Hadoop offering.
Ted also announced a new update to the Microsoft PDW which will be available in the 1st half of 2013 and will be called the SQL Server 2012 Parallel Data Warehouse. This update will include some of the features which will be included in the next major release of SQL Server such as the updatable column store index. Another big enhancement of the new SQL Server 2012 PDW is that SQL Server Management Studio will now be the query tool which is used to run the queries against the PDW system. During the demo against the PDW doing a simple count against a 1PB table the query ran in just a few seconds. Doing data aggregations against the same table for a real world style reporting query the query again ran in just a few seconds.
The next announcement was Polybase. Polybase allows you to query data from an external source from within the SQL Server 2012 PDW. This will allow you to create an external table which simply links to a Hadoop cluster. This will be done with the CREATE EXTERNAL TABLE statement which accepts the schema that you want and a hdfs URL to Hadoop without needing to create a linked server to the Hadoop cluster.
The next step of BI is here in SQL Server 2012 SP1 and Office 2013. Office 2013 includes PowerView natively within the product. It also supports some new mapping features natively within Office 2013 without any additional downloads or plugins. This support includes term extraction which is done through HDInsight. This allows us to view text data and extract names, sentiment, locations, etc. from simple text data and use those extracted strings in order to build PowerView reports. PowerView also now supports DAX queries to a MOLAP cube straight out of the box.
Stay tuned for more updates from the PASS Summit through out the week.