January 13, 2005 at 7:21 pm
I'm in the planning stages of implementing a reporting server. I plan to use transactional replication to replicate a 100gb database and would appreciate help with a few questions.
1. I want to minimize the impact of the initial snapshot as well as the effort to restore the reporting server and resynch in the event I lose the reporting server. Instead of pushing an initial snapshot, is it possible/advisable to start with a copy of the production db and initiate replication as opposed to starting with a snapshot? Any other insights to restoring and resynching would be appreciated. We're considerting retaining more history on the reporting server than we have in production and I'm not sure how to deal with that in the event we lose the report server.
2. I understand that it's advisable to place the distributor on a separate db from the publisher and the subscriber. Any benefit to putting the distributor on the subscriber?
Thanks
January 17, 2005 at 8:00 am
This was removed by the editor as SPAM
January 17, 2005 at 9:14 am
I would use a completely different server as the distributor, it offloads all the work to that server, easing up the load on the other two.
As for the snapshot, you have a couple of choices, you could run a concunrrent snapshot of the table (this does so without the table lock so users can still access the data, although performance will be hindered). Alternatively, you could run a backup of the db/table, set up replication, restore on the other side, and when you add the subscription check the radio button stating that the subscriber already has the schema and data, this will then not snapshot.
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply