July 16, 2003 at 9:11 pm
Any good way to measure the SQL performance when setting up the replication? Is replication taking a lot of resources? How could we tell the impact if I have single replication setup compare with multiple replication setup in the same server? Anyone can share with me?
Thanks in advances.
July 17, 2003 at 5:03 am
MS has a good whitepaper on perf and tuning on MSDN. Overall I'd say the impact is fairly minor, but it's there. Logreader and distribution agents use some memory, only an issue if you have a lot running and don't have a lot of memory. You'll see more disk activity because of distribution db.
Andy
July 17, 2003 at 6:37 pm
Hi,
I has the same concerns. We have a backend database and an app server. Both are single Xeon machines with 2Gb memory.
We used push transactional replication from data server to app server so that, if needed, app server could run the entire application.
Our experience was great (once we got everything going) it had no measurable impact at all. Our system supports an average of 20 trans per second and the average latency between the two databases is around 5 seconds.
Cheers, Peter
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply