What are the applications of distributed database managment system?
Answers
This is a surprisingly tough question to answer.
Distributed databases can mean NoSQL databases or they can mean a relational database expanded out to be distributed. These two classes of distributed databases have somewhat orthogonal requirements.
First, do you want to use a relational database or is your data “Big Data” that does not require all the heavy overhead of a relational algebra model?
I’m guessing that if this is for educational reasons, then you probably want to use a relational distributed database.
Relational databases that are distributed usually either shared-everything, where the replication is done at the hardware level and the database transparently gets deployed across a cluster as if it were a single instance, or shared-nothing, where the database gets sharded into multiple nodes, sometimes with replicas.
Most relational databases these days have some sort of distributed capability.
Shared-everything:
Oracle, DB2, probably some other “big” vendors
Shared-nothing:
MySQL, Teradata, etc.
There are other architectures that are explored in academic literature, but they tend to be more common among NoSQL models which IMO do not implement them to their maximum consistency potential — a shame because I think there are really interesting architectures that could be developed if there were a need for stronger consistency other than “row level.”
Why would you want to use a distributed database? Simple, for replication support (the data persists through hard drive failure, and you want more application control over the backup over what a nightly tape drive can provide) and also good when the database itself exceeds the hardware resources that one machine can provide.