My PASS Summit 2013 Presentations

I had the honor to speak for the first time at this past PASS Summit 2013 in Charlotte, NC. It was a great opportunity for me as this is one of the most prestigious nationwide conferences on SQL Server. I delivered 3 different sessions on Business Intelligence related topics. Attendance to my sessions exceeded my expectations and my overall experience as a first-time speaker was very positive. I also need to add that the God’s of the live demos were on my side.

My session titled PowerPivot + Power View = Self-service BI Revolution was full and had about 5 to 8 people standing in the back. Everybody got excited when I did a demo of the nice interactivity in Power View and 3D geographical capabilities of PowerMap.

The Lightning Talk room was packed and people had to be turned away. This room should be bigger next time as these Lightning Talks are very popular. Everybody accepted and agreed with my postulation on the need for Conformed Stored Procedures in addition to Conformed Dimensions and Facts. After the Lightning Talks were over, several attendees came forward to discuss further how Conformed Stored Procedures should be implemented. One attendee in particular, mentioned that he had been thinking about proposing a similar approach to his team and this talk helped him get re-assurance of this approach.

My last session on SSAS Hierarhcies Tips & Tricks was a half-day session on the last day of the summit between 2:00pm and 5:00pm. It was half-full for the first half of the session, and ended up about 1/4 full.  This drop-off is expected towards the end of the Summit as many people fly back home on Friday and usually leave around 4:00pm.

I’m very pleased with the turnout and feedback received. The audience in all of my sessions was very engaged and had a lot of great questions. I definitely look forward to speak at PASS Summit 2014 in Seattle, WA and at PASS BA Conference 2014 in San Jose, CA.

To download my presentations slides click on the links below:

1. Regular Session (75 minutes): 4798 PowerPivot + Power View = Self-Service BI Revolution
2. Lightning Talk Session (10 minutes): Slides: 5511 Conformed Stored Procedures Script: ConformedStoredProcs_SQLScript
3. Half-Day Session (3 hours): 4803 Fast Performing SSAS Hierarchies Tips and Tricks

Thanks to all the attendees and PASS team for this great opportunity. I hope I met or exceeded everyone’s expectations.

If you attended one of my presentations, please leave your comments and feedback!

.Net Code Camp Orlando 2013

This past Saturday March 16th, I had the opportunity to speak once again the .Net Orlando Code Camp.  I presented a session under the SQL Server track titled “Change Data Capture, a developer’s best unknown friend”.  I had better than expected turn-out with about 20 attendees and overall very good participation.

I was glad to hear from more than one attendee that they were planning to use Change Data Capture in their environment and best of all to get rid of triggers!

I had the opportunity to catch-up with some of my friends and fellow SQL-ers from PASS. Also, met some new folks from the App/Dev community. These type of events are always great to expand your network, share your knowledge and learn something new!

Thanks to the organizers, sponsors, volunteers and attendees to make this event a success and for having me once more.

You can download my PowerPoint presentation and demo files at: http://sdrv.ms/15UUX64

Tuning the Performance of Change Data Capture in SQL Server 2008

Great SQL Server Best Practices Article on performance considerations for Change Data Capture:

http://msdn.microsoft.com/en-us/library/dd266396.aspx

Change Data Capture as a tool for Business Intelligence, Disaster Recovery, Auditing, and more

This is a summary of the Change Data Capture presentation covered in my sessions at SQL Saturday #38 – Jacksonville and SQLSaturday #40 – South Florida.

Presentation Summary
Change Data Capture (CDC) is one of the new features available for SQL Server 2008 Enterprise and Developer Editions, which allows you to easily track and keep history of all DML (Data Manipulation Language) changes on user tables such as INSERTs, DELETEs, and UPDATEs with little or no overhead. Change Data Capture has been used primarily to track changes for incremental loads using SSIS for the ETL process in Data Warehousing and Business Intelligence because it allows you to identify new data and data that changed or was deleted since last incremental load.

Before Change Data Capture in SQL Server 2005 and prior versions, identifying new or changed data required some level of custom programming, triggers or third party tools. These solutions work well but introduce additional overhead to your production systems. This is the main reason why Change Data Capture was introduced in SQL Server 2008; to eliminate overhead while tracking new and changed data.

But Change Data Capture use is not limited as a tool for ETL Process. It can be used for many other purposes where overhead, simplicity and cost might impose a restriction.  Some of these scenarios include:

1) You are a developer adding or modifying existing code and you need to see the difference in the data output on a particular table.
2) You are a DBA and need to audit several tables in a database and need to report which data was changed, what were the new and previous values, who did the change and when was the change made.
3) You were tasked to delete several records on a highly transactional production database and unintentionally deleted records that you were not supposed to with no possibility of rolling back the changes (no explicit transaction) and no possibility of restoring a backup.
4) You are a DBA and need to know which tables and columns are being written to the most in your database.
5) You are a DBA and need to know how many new records are being entered into your database in a period of time and the percentage of change of existing records.

In most cases you would want to know what changed and which were the values before the change.

 One of the most important benefits of this feature is that it allows you to see the data before & after an UPDATE or DELETE statement, which in turn allows you to query and recover data overwritten or deleted quickly without the need of a database restore. It also captures each new row inserted.

CDC accomplishes this task by reading the committed operations from the log file and inserting the changed records in a tracking table that mirrors the source (tracked) table. The records inserted in this tracking table contain both the value before the change and the value after the change along with the metadata associated with the change. The metadata can be queried to identify the type of DML operation as follows:

1 = Delete
2 = Insert
3 = Update (record’s value before update)
4 = Update (record’s value after update)

More concepts are covered in my PowerPoint presentation which you can download here or by clicking slide below.