This time of the year, a common question that I am asked not only at User Group Meetings, but over email and such is why I am such a big fan of the PASS Summit and why I believe this is a must attend event for everyone. This year I even have the opportunity to have an answer posted on the TechNet SQL Server Blog, that I think is being posted later this week. There are so many reasons as why to attend PASS, but a big one for me is the opportunity to have visibility into SQL Server features that I haven’t had the opportunity to work with. If you are at all like me, my time researching is spent looking at things that will make my current or known future projects even better. I am not spending time like I should learning about features that don’t impact me today, I would love to but I just don’t have all that time.
One of the really neat things about the PASS summit is the ability to attend sessions, make notes on features that I use for reference at a later date, when that information is more relevant to the projects I am working on. Here is a great example; I am really looking forward to attending this session by David DeWitt. A great speaker with a lot of good information, this session will give me some great insight that I might be able to use in the near future.
During my “Big Data” keynote at the 2011 PASS conference, I introduced the concept of an “enterprise data manager” – a new class of database systems capable of executing queries against both traditional structured data stored in a relational DBMS and unstructured data stored in HDFS, Hadoop’s distributed file system. In this talk I will describe the progress we have been making on this concept at the Gray Systems Lab. Extending the Query Processor in SQL Server Parallel Data Warehouse by adding a new table distribution type for data stored in HDFS, we give it the ability to query data stored in HDFS without first having to load the data into PDW. By leveraging a database QP, we can perform real optimization – transform selections, projections, aggregates and other operations on HDFS files into MapReduce jobs and execute them on the Hadoop cluster as part of the query execution plan it generates.
So anyway, this is just one other reason why I really believe the PASS Summit is a can’t miss event. I hope to see you there.