Webinar Q&A: Getting Started with Change Data Capture
Bradley Schacht is a Principal Program Manager on the Microsoft Fabric product team based in Saint Augustine, FL. Bradley is a former consultant, trainer, and has coauthored 6 SQL Server and Power BI books, most recently Learn Microsoft Fabric. As a member of the Microsoft Fabric product team, Bradley works directly with customers to solve some of their most complex data problems and helps shape the future of Microsoft Fabric. Bradley gives back to the community through speaking at events such as the SQLBits, Fabric Community Conference, PASS Community Data Summit, SQL Saturdays, Code Camps, and user groups across the country including locally at the Jacksonville SQL Server User Group (JSSUG). He is a contributor on SQLServerCentral.com and blogs on his personal site, BradleySchacht.com.
Q: Does disabling CDC remove any objects that were added? A: Yes, when you disable CDC on an individual table the associated change tracking table as well as the auto-generated functions are removed. If you disable CDC on the database all change tracking and auto-generated functions are removed. The wrapper functions are not removed automatically, you must remove those yourself. Q: How does the log reader job impact the transactions being written in the application? A: The log reader only does reads on the transaction log. It does not lock the log or anything like that. There should be no affect on the applications writing to the tables. Q: Will CDC work with Simple recovery mode? A: Yes, CDC works with both full and simple recovery modes. Q: Can CDC be enabled on MDS (Master Data Services) databases? A: CDC can be enabled on any database as it just reads the logs. That being said you do need a primary key or unique index on the tables. I'm not sure I see the value of putting CDC on an MDS database though. Those tables are setup such that you need to go through the views to get data back. Q: Does CDC work on Hecaton databases? A: I don't see why not, although I will need to test this as I have not tried it. Ultimately there is still the transaction log involved as far as I know so I would venture to guess yes. Don't at all hold me to that as I haven't done any real world work with Hecaton so testing and a follow up blog will be required there. You've definitely peaked my interest there though. Q: What do you recommend for getting the first pull of data (initial load), date or LSN based queries? A: I prefer to go the date base route using the control table. The first time you pull data pass in NULL for the begin date and you will get back all of the data from the beginning of time. Quick and easy. Q: Is CDC enabled on the source system or the destination system (Data Warehouse)? A: CDC will be enabled on the source system. You want to leverage the deltas on the source side so you have less data to load into the warehouse and less complex logic to do on the warehouse load ETL. Q: Can you enable CDC on a view? A: No, CDC can only be enabled on tables. Q: What version of SQL Server is required? A: An Enterprise edition of SQL Server is required to use CDC. The feature is available in 2008 or later. I also did a similar presentation last night for the Orlando SQL Server Usergroup (OPASS, website) on CDC with SSIS. That focused on using Change Data Capture with the new SSIS 2012 components. I have consolidated the slide decks and included both SSIS projects in the download link which can be found in the giant information box below so no one can miss it. Some notes on what is included: [important]