Click here to monitor SSC
SQLServerCentral is supported by Red Gate Software Ltd.
 
Log in  ::  Register  ::  Not logged in
 
 
 
        
Home       Members    Calendar    Who's On


Add to briefcase

ETL testing Expand / Collapse
Author
Message
Posted Wednesday, October 24, 2012 11:43 AM
SSC Rookie

SSC RookieSSC RookieSSC RookieSSC RookieSSC RookieSSC RookieSSC RookieSSC Rookie

Group: General Forum Members
Last Login: Tuesday, November 20, 2012 9:04 PM
Points: 28, Visits: 86
Hi All,

I want to know if there is any automated tool which can be used to for post migration ETL testing? The ETL has been written using sql scripts to move the data from Legacy database to a new database and both have different database structures. We need to test the tranformed data for accuracy and check for data quality and quantity after the ETL process is executed.

I have questions about the following:
1.what should be the best approach/ best practices to test the data in the new database after transformation?
2. I would like to know if there is any tool which can be used for post migration ETL testing and make the tetsing process faster? Thanks
Post #1376607
Posted Thursday, October 25, 2012 8:35 AM
Old Hand

Old HandOld HandOld HandOld HandOld HandOld HandOld HandOld Hand

Group: General Forum Members
Last Login: Saturday, January 25, 2014 2:15 PM
Points: 317, Visits: 3,612
vick12 (10/24/2012)
Hi All,

I want to know if there is any automated tool which can be used to for post migration ETL testing? The ETL has been written using sql scripts to move the data from Legacy database to a new database and both have different database structures. We need to test the tranformed data for accuracy and check for data quality and quantity after the ETL process is executed.

I have questions about the following:
1.what should be the best approach/ best practices to test the data in the new database after transformation?
2. I would like to know if there is any tool which can be used for post migration ETL testing and make the tetsing process faster? Thanks


Your questions are very valid, but at the same time also difficult to answer...

It all really depends on the nature of your system. For instance, if you are importing order data it could be as simple as doing a count of orders and looking at order totals of a given time period. If you are dealing with certain types of analytical systems however it could be a lot more complex than that, because a simple record count and total may not provide enough information to validate the data.

To answer your questions though:
1. The best approach is typically one that works in your environment. If you know that x amount of records in your staging area needs to amount to x amount of fact records, a simple comparison could be enough.

The most important aspect is to define your criteria upfront. This would be a collaborative exercise between the business and technical areas...you may be interested in record counts, but the business may want to see total orders by region (as an example) to feel comfortable with the data. It is very important that both and and your business users have confidence in the process, and you will have to agree on the parameters that provide that confidence.

2. I don't know of any such automated tool. There should be some tools which could help you profile the data, but you can do that with the Profile Data task in SSIS as well. Spend some time and develop a data validation process. It should only have to be done once, and will be invaluable down the road.

Hope this helps.



Post #1377034
« Prev Topic | Next Topic »

Add to briefcase

Permissions Expand / Collapse