Database Management System

Spread the love

Transaction-Level Consistency

The transaction-level consistency, on the other hand, offers the indication that the set of rows is going to remain consistent all through the dispensation of all statements that are within the transaction. When dealing with a single-user database, the user enjoys the freedom of being able to modify the data in the database with no unease for other users adjusting the same data at the same time. On the other hand, the case of a multiuser database follows that the statements that are within manifold concurrent transactions can renew the same data (Sippu&Soisalon-Soininen, 2015). The transactions that are executing at the same time should produce meaningful along with consistent results. As a result, control of data concurrency along with data consistency is fundamental in a multiuser database.

In describing the consistent transaction behavior at what time transactions are running at the same time, database researchers consequently come up with a defined a transaction isolation model referred to as serializability. The serializable approach of a transaction behavior makes attempts to ensure that t transactions are run in such a manner that they come out as being executed one at a time, or serially, instead of concurrently. Although the degree of isolation linking transactions is normally pleasing, running numerous applications in this mode can critically impact application throughput (Sippu&Soisalon-Soininen, 2015). Absolute separation of concomitantly running transactions could imply that one transaction cannot implement an insert into a table being inquired by another transaction.

When implementing a transaction-level read consistency, transactions runs in serializable mode, with all the data access reflecting the condition of the database as of the time the transaction started. The implication of this attribute is that the data that is seen by all queries inside the same transaction is constant on a unitary point in time; except for those queries executed by a serializable transaction see the changes made by the transaction (Korotkevitch, 2014). Transaction-level read consistency leads to the production of repeatable reads moreover does not produce a query to phantoms.

Statement-level Consistency

The case for statement-level consistency encompasses the assertion that a statement applies to a set of rows that is constant from the start of the statement until the conclusion of the statement. The use of Oracle results to the constant enforcement of statement-level read consistency. The implication of this attribute is that guarantees that all the data that is returned by a unitary query is from a solitary point in time, with the core reference being to the time that the query started. Consequently, a query by no means sees dirty data otherwise any of the modifications that are made by the transactions that entrust during the execution of the query (Korotkevitch, 2014). Just as the execution of the query proceeds, it is only then data that is committed before the query began that is visible to the query. In this case, the query does not assess any changes that could have been committed following the statement execution begins.

The consistent result set is offered for each query, an attribute that guarantees data consistency, with the user not required to carry out any action. The SQL statements encompassing the SELECT, INSERT with a subquery, UPDATE, along with DELETE all query data, also overtly or unreservedly, return consistent data (Korotkevitch, 2014).The SELECT statement is normally an overt query and can possess nested queries or join operation. The INSERT statement can employ the nested queries. UPDATE along with DELETE statements can utilize WHERE clauses or subqueries in affecting only a few rows in a table instead of all rows.Queries employed in INSERT, UPDATE, along with DELETE statements are certain of a consistent number of results. Nevertheless, they do not assess the variations that are made by the DML statement itself, which implies that the query in the operations assess data as it existed earlier than the operation started to make changes (Korotkevitch, 2014).


Korotkevitch, D., (2014). Pro SQL Server Internals. Apress

Sippu, S., &Soisalon-Soininen, E. (2015). Transaction Processing: Management of the Logical Database. Springer.

© 2021:, All Rights Reserved | Innovation Theme by: D5 Creation | Powered by: WordPress
error: Content is protected !!