0

We are replicating a very large database to a data warehouse server (DELETE is off) via publisher/subscriber. The data warehouse team want to change some of the data at their subscriber side by wiping to NULL a few non-key data columns (name, address etc.).

I have tried searching via Google etc. but can't find any information on whether this is possible without breaking replication or causing unknown issues.

We do NOT want these changes propagated back to the publisher.

Anybody know what impact this will have?

I could change the Sp_MS* stored procedures but I reckon on reinitialize of the replication they could be overwritten by the MS standard ones again – I don’t really want to do this as it seems a messy solution.

I can test the idea out but was more concerned even if it did work with no issues that some unknown factor would cause an issue at some point.

Skyline
  • 103
  • 2
  • 6
  • For Tx replication, I wonder if you could define a column filter on the articles to exclude columns you don't want to include in replication. When the tables are created on the subscriber the defaults would be null, you would have to be sure to exclude check constraints for default values for replication. – Ross Bush May 04 '18 at 15:17
  • Thanks Ross, we had looked at that idea but lots of not null columns we were unaware of unfortunately so wiping to NULL is out too. – Skyline May 07 '18 at 17:31

0 Answers0