How do I go about stopping transactional replication on the Subscriber page?
We have a publisher on server A, a distributor on server B, and a subscriber on server C. All are on the same network.
Background: We have a rather busy BizTalk server where we replicate the data from the BAM parts to a separate database server that is used for a custom web application for statistics, searches, and so on. The tables on the BizTalk server are both fairly large and missing indexes, and so on. Because they are basically just there to pull data into memory, we've opted for the second, replicated database, in which we have several optimizations to read the data.
One thing I noticed, though, is that adding a simple, nonclustered index takes an infinite amount of time and causes all sorts of locks (we use SQL Server Standard, so online indexing is not an option) database some of the time Writing from the BizTalk data while creating the indexes would be a good idea (right or wrong?).
The question is, how can I stop replicating at the Subscriber (without losing publisher data) while creating the indexes?