October 22, 2004 at 8:51 am
Ok I’m not an SQL wizard, all I want to do is the set up transactional replication.
I have a new Win2003 server just for SQL 2000 I want to use the old Win2000 SQL Server as a backup of the main (new sever) it's in a different building running over a 1GHz Fibre backbone.
So What is the Publisher, what is the distributor and what is the subscriber, and what is Pull ?
I’ve read half a dozen articles and I am still confused.
Surely this should be easy with the built in wizards?
Guidance would be appreciated.
Scott Jordan
October 25, 2004 at 8:00 am
This was removed by the editor as SPAM
October 25, 2004 at 10:55 am
Based on my (limited) knowledge of replication here are a few answers:
Publisher: the server that actually has the data
Distributor: server that makes the data available (we have our publisher configured as the distributor as well)
Subscriber: the server that is updated with fresh data from the publisher
push subscription: updates are initiated (pushed) by the data source
pull subscription: updates are requested (pulled) by the subscription server
the wizards are pretty helpful ...
open enterprise manager
rt-click on replication
select configure publishing, subscribers, and distribution ...
follow the wizard
then rt-click on publications (on the publication server) and select 'new publication'
do the same for subscription
hope this helps (if i am wrong ... please correct me)
October 26, 2004 at 2:06 am
Thanks Seth
I've been doing a bit more reading and have now discovered that for transactional replication to work every table replicated has to have a unique ID column, problem is this is not my data so i can't add anything.
I'm looking at third party solutions now, but some seem very expensive.
I see SQL 2005 has a new database "mirror" option, but can't realy wait for 2005 to go gold.
Cheers
October 29, 2004 at 5:23 pm
Scott,
You should look at log shipping solutions. These are very simple to set up and do a good job of keeping databases in sync within a specified time interval. Log shipping scenarios can be scripted or it can be set up given you have the enterprise edition of sql server.
The log shipping process goes as follows:
1) You have a source and a target server
2) Take a full backup of the source and restore it onto the target
3) Take log backups of the source every 5 minutes or so and ship them to the target to be restored.
I have been using this process and it is very simple and works well. If you would like any additional information let me know.
Jason Hall
November 21, 2004 at 12:40 am
You can use "snapshot" replication instead of "transactional" replication for tables that don't have a primary key. This will replace the table each time instead of just sending changes, so it's best for smaller tables
Viewing 6 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply