Re: Transactional Replication
From: Paul Ibison (Paul.Ibison_at_Pygmalion.Com)
Date: 03/01/05
- Next message: Paul Ibison: "Re: Identity error in replication"
- Previous message: Kristy: "high latency with Dist agent"
- In reply to: Kyle Jedrusiak: "Re: Transactional Replication"
- Messages sorted by: [ date ] [ thread ]
Date: Tue, 1 Mar 2005 21:29:13 -0000
Kyle,
my article is not intended to be a how-to, but more of a high-level
overview, which sounds as though it's still relevant here: BTW, for full
details of high-availability solutions, have a look at the HA articles on
the MS website. Anyway, in my article I mention the degree of latency
involved. For very fast latency and automatic failover there is clustering,
but the distance between nodes looks as though this is not possible in your
scenario. Transactional replication will offer the next fastest latency, but
will alter the schema, and finally log shipping will not alter the schema,
but will increase the latency. My preference would be for transactional
replication with queued updating subscribers, where the subscriber would be
immediately useable afterwards.
Can you just change the IP address/DNS entry of your client app after
failure? Yes this is usually possible. However, this usually means that the
servers have different names, and any 4-part names will fail if you're using
linked servers.
Rgds,
Paul Ibison SQL Server MVP, www.replicationanswers.com
(recommended sql server 2000 replication book:
http://www.nwsu.com/0974973602p.html)
- Next message: Paul Ibison: "Re: Identity error in replication"
- Previous message: Kristy: "high latency with Dist agent"
- In reply to: Kyle Jedrusiak: "Re: Transactional Replication"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
