What is Replication?
Replication is another way of disaster recovery with the help of this we can recover our database and we do not need to do much manual efforts.
Replication is the way by which you can transfer data or database object from one server to another server, synchronize them, and maintain them identical.
The interesting part is you can implement replication for any location like LAN, WAN, Internet without much effort.
What is the initial requirement for Replication?
The replication can work on SQL server 2000 2005, 2008.
What are the good points of Replication?
There are many advantage of replication
- High availability solution
- Minimum error
- Easy to implement
What are the points which we would not like for replication?
- Main server overhead increase
- If there is failover occurs then we have to handle that manually
What are the terminologies we used in Replication?
We used following terminologies in replication
The publisher is main database source which publish data or other data object to different server or we can say subscriber.
2) Subscriber :-
The Subscriber by name it is clear that it subscribe the database from publisher server.
3) Distributor :-
The distribution is the server which manages flow of the data. It not mandatory to have distributor always. A publisher sever can manage role of both (publisher & distributor).
An article can be any data object like data table, Rows, Store procedure or view. A publisher can be published number of article.
What are the different types of Replication?
We can implement 3 type of replication
Snapshot Replication, Transactional Replication, Merge Replication.
By the name it is clear that in this publisher take snap shot of entire database and publish to the subscriber. And when its take full snap shot obviously it taking it consumes more resource. Or we can say its resource intensive process. It also known as simple replication. It used where database changes are rarely.
Transactional Replication: It also known as dynamic replication.
In this the publisher and subscriber are always synchronized. It used where we need exact same copy on subscriber with immediate effect. It can be useful when subset of database or table is required.
It’s very interesting replication by the name it’s clear that it use merge method. In this both publisher and subscriber has freedom to work without network and when they connect the replication agents check both the database and merge each other’s changes. In merging if any conflict occurs than preference is given to publisher.
Later on in the post we will get a brief introduction how to implement replication tillthan enjoy definition :).
Thanks & Regards