Distributed database
-> Data in distributed database system is stored across serveral
Properties
-> Distributed data independence :Ask Queries without specifying where referenced
relations
are located
-> Distributed transaction atomicity: Write transcations that access and update
data at
several sites
Type of Distributed Database
-> Homogenous Distributed Database System: If data is distributed, but all server
run the same Dbms Software
-> Heterogenous Distributed Database Systems:It is a system where different sites
use different dbms but ar connected to access to data from various sites
ARCHITECTURE
->Client -server systems:
* This systems has one or more client process and server process
*A client process can send a query to any one server process
*Client are responsiable for user interfaces
*Server manage data and execute trasactions
*Client process can run on a personal computer and send queries to server running
mainframe.
*Architectures not allow single query to span multiple servers
->Collaborating Server Systems
* The client process would have to be capable of breaking of query into subquery
* The collaborating server system can have a collections of database servers each
capable of running trasactions against local data
* When a server receives a query that requires access to data at other server ,It
generates appropriate subqueries to be executed by other servers
*Puts the results together to compute answers to the original query
-> Middleware systems
*Middleware is and intermediate layer of software that sits b.w applications and
network.
*Message -oriented middleware is designed for the purpose of transporting message
between two or more applications and best suited for distributed applications thath
requires transactions
Types of fragmentation:
->Horizantal fragmentation:The union of the horizantal fragments must be equal to
original relations. Fragments are disjoint
->Vertical Fragmentation : The collection of vertical fragments should be
lossless-join decomposition.
