Transparencies stands for separation of high level semantics of system from low-level implementation issues. High level semantics stands for endpoint user, & low level concerns with complicated hardware implementation of data.
Types of transparencies:-
- N/W /Location: - DBMS users should not be concerned about the type of DBMS they are using, the network should not be visible to the user. Thus to provide n/w transparency, we need following:-
- Query language that will be should not include any location specification, in this way, location transparency can be achieved.
- The data that’s stored in a relation should not contain any location specific name so that naming transparency can be assured.
- Every database object must have a system wide unique name.
Using aliases, we can move database objects transparency.
Replication Transparency: - It states that the replicate created should be controlled by system not by user.
- The user should not have any doubt whether the featured data is coming from replicated copy of relation or from actual copy of relation.
To achieve replication transparency, concurrency control protocol needs to be used which assures that update of data taking place in one copy should also be updated in other copies. In this way, transparencies regarding replicas of data can be maintained.
Fragmentation Transparency: - It states that fragments are created to store the data in distributed manner should remain transparent & all data management work required to control fragments should be done by system, not by user. In this task, when a user puts a query, global query is distributed in many sites to fetch data from fragments & this data is put together at end to generate results. The system ensures that the total procedure of query decomposition & re-composition should be transparent to the user.
and 5 others joined a min ago.
teamques10
★ 68k