Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

peer 2 peer replication

0 views
Skip to first unread message

Larry

unread,
Jun 25, 2001, 6:29:13 PM6/25/01
to
We am looking for the best replication solution for our needs:
- Data will be in two cities (San Francisco and Chicago)
- Data changes fairly rapidly (~500 inserts per day)
- Changes may occur at either location
- Synchronization latency must be minimal (prefer 10 sec or less, could
live with 1 min)
- VPN over the internet is the conduit for data between the sites
- Data is about 1 gig in about 50 files

Our current thinking:
- We will not use clustering
- Each city will have an active and a standby SQL server
- Merge replication will be used to distribute data
- The active SQL server in each city will use push subscription to the
standby
- The active SQL server in one city will use push subscription to the
active SQL server in the other city

Any comments?

Thanks,
Larry

Jason Burjaw

unread,
Jun 26, 2001, 9:47:56 AM6/26/01
to
The best P2P replication solution is PeerDirect Replication Engine
(PDRE). PDRE can provide you maximum reliability, the most efficiency
of any replication product on the market, the ability to allow updates
at any site (not just at a single "master" or "publisher" site),
seamless cross-platform capability (we support all major database
platforms, incl. MS SQL, Oracle, Sybase, Informix, as well as smaller
"desktop" databases such as Paradox, Access, InterBase, Pervasive).
Not only is PDRE very quick to set up, but it requires the least
administration, and provides the greatest return on investment of any
replication product. You can find out more at
"http://www.peerdirect.com"

Jason


"Larry" <LarryD...@hotmail.com> wrote in message news:<euqwDZc$AHA.1108@tkmsftngp05>...

Scott Whigham

unread,
Jun 27, 2001, 10:53:47 AM6/27/01
to
Sounds like you have a plan and that it will work. You don't need to use a
third party tool, IMO - the reason you pay all the $$$$ for SQL Server 2000
is so that you get those tools....

I am curious as to why you are choosing not to use clustering and are,
instead, choosing replication. This will cause a load on your distributor
(which sounds like your publisher). Your choice is not all that uncommon
but I always wonder if it is the most effective.

A key thing to take into account is what you will publish - does your front
end use stored procedures to make changes to the data or is it embedded SQL?
Big difference in how you might choose to implement replication.

If you'll post any replies, maybe we can help more.

HTH

Scott Whigham

"Larry" <LarryD...@hotmail.com> wrote in message
news:euqwDZc$AHA.1108@tkmsftngp05...

0 new messages