Posted on Leave a comment

Calling SQL gurus – Is there a way to programmaticly prove data accuracy?

Say I have an existing MS SQL database with many tables. Now I’m going to import many more records. From a quality assurance standpoint, I’d like to be able to confirm that the existing data prior to the import was not inadvertently altered during the import of the new data (since the import includes some manipulation of the new data).

Is there a magical way to do this. I’m stuck on comparing record counts but that does really tell me if the existing data was altered during the import process.

Conceptually, I’d like to be able to essentially do something like a hash of the complete dataset before the import. Then after the import, query for all the old records, rehash and compare the two hashes.

Thoughts?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.