Groups | Blog | Home
all groups > sql server replication > november 2005 >

sql server replication : Sync Check


CLM
11/21/2005 3:30:04 PM
I'm about to implement transactional replication from 2000 to 2000 for a
"failover" type of situation. That is, if we lose the publishing server, we
can pretty easily point various apps, etc. to the subscribing server. But
that leads to a couple of questions if such an event actually occurs (where
the publisher goes down for an extended outage):

1. Now I want to repoint to the subscribing server. I look at my
Replication Monitor and all looks well - everything is bright green. But how
do I really know the tables are synced if this has been running for two
months uninterrupted? I know Informix has a tool that will compare pubisher
and subscriber - does Sql Server have something like this? What is the
standard way of handling this situation?
2. If the publisher does go down (and I have access to Enterprise Manager,
etc.) and now I'm point the apps to the subscribing server and the app is
making changes there - then is there some "slick" way to resync in the
reverse direction? (Besides putting replication on the subscribing server
instead or just doing a DTS, etc.) Again, I can think of several ways to do
this (after shutting down the apps) but just wanted to see if there was a
clever way to do this most easily and accurately?

Hilary Cotter
11/21/2005 10:13:05 PM
1) Run a validation. This will confirm everything is the same.
2) No, what you have to do is use merge replication which will automagically
do what you want, or use bi-directional transactional replication and have
different identity seeds on either side, i.e. odd numbers on the publisher
and even on the subscriber and use an increment of 2.

--
Hilary Cotter
Looking for a SQL Server replication book?
http://www.nwsu.com/0974973602.html

Looking for a FAQ on Indexing Services/SQL FTS
http://www.indexserverfaq.com

[quoted text, click to view]

CLM
11/22/2005 2:04:02 PM
A few questions for you about the bi-directional set up (because there are
many, many identity columns):
1. What do you mean by bi-directional? Are you saying that I can have, for
example, transactional replication set up from serverA.tableA to
serverB.tableA and then have a separate transactional set up from
serverB.tableA to serverA.tableA?
2. I thought that you had to turn the identity column off on the target
server, i.e. if you had transactional replication from serverA.tableA to
serverB.tableA that serverB.tableA had to have the identity columns turned
off? If that's true, then how can you have this set up realtime?
Thx so much for the help...

[quoted text, click to view]
AddThis Social Bookmark Button