Analysis of Transaction and Concurrency Mechanism in Two Way Waiting Algorithm for different Databases

Database is the well-organized collection of data in a meaningful way that user can perform the transaction. Transaction is a sequence of many actions considered to be atomic. In the coming months or years you will probably be implementing applications for multiple users to use across a network. Client/server environments are designed specifically for this purpose. Traditionally, a server (in this case, a database server) supports multiple network connections to it. As often happens with technology, this newfound flexibility adds a new degree of complexity to the environment. The throughput of the transactions can be increased by performing the transactions in parallel. Transactions are performed in parallel without infringing the data integrity. When the transactions are executing in parallel then this will lead to concurrency in database and deadlock of the transaction. This work will analyze how and is done to prevent the occurrence of deadlock and also improves system performance by shrink the number of restarts transaction which also saves the time and cost for performing the transaction in the system. So, it improves the efficiency of the distributed system. Keywordsconcurrency; transaction; timestamp; lock; Conflict (serializability, precedence);Background Process in Oracle; Commitment; consistency ; deadlock; direction.