"Simon" <simoncar@NotThisBittotalise.co.uk> wrote in message
news:urOZtGbRFHA.3788@tk2msftngp13.phx.gbl...
>I am in the early stages of setting up replication and I was wondering if
>someone could offer a little advice.
>
> I have an updateable database of around 200,000 records, each described
> (in great detail) across a few hundred tables. There are also associated
> tables such as decode tables. The data changes by about 1% a day.
>
> I want this data for manipulation and publishing on a web site, not
> updateable, available 24x7 - not 100% availability but just minimising
> data unavailability whilst data is refreshed.
>
> I have set up transactional replication. I have then set up a pull
> subscription from the subscriber. When I first set it to reinitialise and
> then run the snapshot and distribution agent the tables are created
> successfully. Every day (or every 10 minutes in my test environment)
> transactions are pulled down and applied to the replicated set.
>
> I then tested by adding a new table (article) to the publication. It
> seemed that the only way to get the intial schema and data down was to
> resynchronise the snapshot. Unfortunately this meant the whole replicated
> set was re-created. Is there a way to propagate a new (or structurally
> different) article to the subscriber?
>
> Cheers!
> Simon
>