what is parallel system in database
It is a database system based on parallel systems which consist of many memories and processors and they are connected by network cabinet. It likely to exploit modern software oriented for data management. Conventional DBMS has a problem of high disk access time. They are very large data bases that cannot be used by single systems. To cope with problem we use parallel data base systems that store huge databases.
Advantages Parallel Database System
Advantages: query parallelism is increased and the response time is decreased. It is highly available because it uses data replication and extensibility of data.
Sharing:
– Memory:very simple and load is balanced, but cost is high and less available.
– Disk: low cost, load is balanced, easily available, complexity is higher, and have coherence problem.
– Nothing: then availability will be easy and less cost, problem will be complex, if new nodes are added then they require recognizing data bases.
Database need to be spared across the system for efficient parallel I/O, this technique is known as data allocation. Partitioning of data include three techniques: round robin, hashing, range partitioning. Fragmentation is data is required to increase the system performance and minimize the response time, which will increase the efficiency of output. But it has one problem, it lead to non-uniform partitioning.
Parallel query: when multiple queries are generated for transaction then it is inter query parallelism. Whereas intra query are of two types: inter operator: when we operate several query on several processors parallel then it is said to inter operator. If same query is executed by different processor then it is intra operator.
For parallel joining we need to select A,B and multiply A.x and B.y and then match A.x with B.y and then insert them into the C.
Download Technical Seminar Topic on Parallel System Database .