Post has been published on Varinder Sandhu 's Blog
Post has been published on http://www.varindersandhu.in/2011/05/14/understanding-of-replication/
SQL Server – Understanding of Replication
The primary purpose of replication is to distribute data from a master database to one or more secondary databases. Because Replication maintains a duplicate copy of data in synchronization with the master copy, the technology can be used to provide availability for application.
Snapshot Replication takes the entire set of data and sends it during the each cycle of replication. This is full copy of data that is sent by Publisher to subscriber.
Transactional Replication
Users obtain complete initial copies of the database and then obtain periodic updates as data changes.
We can configure transactional replication with two optional modes:
Immediate updating subscriber
Queued updating subscriber
Transactional Replication can be deployed in two alterative architectures:
Bidirectional transactional replication
Peer-to-peer transactional replication
<!-- google_ad_client = "pub-2404605494811633"; google_alternate_color = "FFFFFF"; google_ad_width = 468; google_ad_height = 60; google_ad_format = "468x60_as"; google_ad_type = "text_image"; google_ad_channel =""; google_color_border = "FFFFFF"; google_color_link = "0000FF"; google_color_bg = "FFFFFF"; google_color_text = "000000"; google_color_url = "008000"; google_ui_features = "rc:6"; //-->
In merge replication received the data from two or more publishers is combined into a one subscriber.
Data conflicts can occurs within any environment that enable distributed processing of transaction.
Three type of conflicts can occurs
The insertion of a duplicate prinmary key on both publisher and subscriber
A conflicting update same row at both publisher and subscriber
An update of a nonexistent row
http://msdn.microsoft.com/en-us/library/ms151198.aspx
http://msdn.microsoft.com/en-us/library/ms151215.aspx