after spending days in search about the best practice to make bi-directional replication between 2 databases on 2 distributed sql server instances ,,i have found that sql server service broker is the most suitbale with my business req to use as i have a
problem with the updatable subscrption replication due to leased line between networks in 1 MEGA and it will be large amount data that will be replicated and that will not work as the latency will increase ,,but i need some help to know how can i resolve these
conflicts when i am using sql server broker service:
1-If you insert a record that has a key into a table on one of the servers and another record that has the same key already exists on the other servers that participate in the replication, the replication does not propagate the changes to the other servers.
2-When you update a column in a record that is updated at the same time on another server, the data may be different on the two servers.
3-When you update different columns in a record, simultaneous updates of different columns of a record may sometimes lead to conflicts.
4-When you delete a row that is being deleted at the same time on another server that is participating in the replication, the replication fails because the DELETE statement does not affect any rows on some of the subscribers.
Hello , You can have merge replication using web sync, you can read about replication options at http://msdn.microsoft.com/en-us/library/ms151198.aspx. Can you explain what
hi jayaprakashi cant use merge replciation ,,as i have to construct messages that contains the values of updated columns only not all the row because i have network with leased line 1 mega ,,so i need to reduce the size of data as much as i canthnx for ur post
Hello , You try column filtering in the article. Thank You Jayaprakash JO – MSFT