Transaction computer

This ensures data consistency between the two accounts

In computer science, and particularly in databases, a transaction such as a booking, a purchase or a payment is implemented through a sequence of operations which move the database to a state prior to the transaction to state B, posterior and mechanisms used to obtain that this sequence is both atomic, consistent, isolated and durable (ACID)The majority of the database management system hierarchical data as relational market, allow people to make transactions atomic, consistent, isolated, and durable. The concept of transaction is based on the notion of synchronization point (sync point), which represents a stable state of the computer system considered, in particular, of its data. For example, when an information operation to transfer money from one bank account to another bank account, there are a task of withdrawal of funds from the source account and a deposit on the target account. The computer program that performs this transaction will ensure that the two operations can be performed without error, and in this case, the change will become effective on the two accounts. If this is not the case the operation is cancelled Both accounts retain their initial values. This ancient technique practiced with the monitors transactional, such as CICS from IBM, TOS BULL, MTU, Siemens, is today widely used in architectures of web applications, and client-server applications. The problem in this mode of operation is that it is sometimes a sequence of several screens or pages to develop a complete transaction ACID.

That is the whole point of the pseudo-transactional

This is the methodology Merise, which has for the first time, defined these concepts: This task is considered to be a pseudo-transaction of a point of view of the monitor is a transaction technique, but of course not really functional as long as the sequence is not completed. The responses of the former are also those which are used today in the 'new' technologies: It is easy to understand why if we had some system locks (DBMS) for the whole sequence, the duration of which is uncontrollable, the system would collapse. But the strategy of the control of isolation is basically functional. The pseudo-transaction is so much ACID, but the functional rules are such that the coherence between each pseudo-transaction of a sequence is guaranteed by the absence of an update of the database. A client application server is well designed, it also uses pseudo-transactions, but the context is managed in the client application, which relieves all the server. The schema type is the following.