Distributed Databases (was Replication Server)

From: Paul Gillingwater <paul_at_eunet.co.at>
Date: 4 Dec 1994 19:52:48 GMT
Message-ID: <3bt6mh$eh4_at_c2.eunet.co.at>


Jeff Pekar (jpekar_at_blkbox.com) wrote:
: Has anyone reading this used replication server for reporting or backup
: puposes? If so, could you contact me at jpekar_at_blkbox.com and let me
: know of your experience with it. Specifically, did you having problems
: keeping your databases in sync, were there any unforeseen problems, etc.

I haven't used Replication Server, but am writing a paper which I am researching now on the different mechanisms for distributed databases. As far as I can tell, Sybase Replication server uses "log sniffing", while Oracle bases its replication upon transaction detection and triggers. I would wlecome comments on the differences and similarities, and some of the following questions:

  1. Which replication mechanism is superior, and what are the trade-offs?
  2. How does each method solve the problem of roll-backs?
  3. Can replication be made to work in a heterogenous environment?
  4. Should replication occur within the database or in the app?

I'd be happy to post of draft of my paper if anyone is mildly interested.

--
paul_at_actrix.co.at (Paul Gillingwater) :: Home Office in Vienna, Austria
If you read news with rn or trn, try EEP! the friendly .newsrc editor.
Waffle users, try <A HREF="ftp://eunet.co.at/pub/news/eep/">WinBoxer</A>
<A HREF="http://actrix.gen.nz/users/paul/plg.html">Visit my URL!</A>
Received on Sun Dec 04 1994 - 20:52:48 CET

Original text of this message