Question

How do optimistic concurrency control techniques differ from other concurrency control techniques? Why are they also called validation or certification tech. niques? Discuss the typical phases of an optimistic concurrency control method.

Short answer

Optimistic Concurrency Control techniques differ from others as they do not use locks to prevent users from modifying data. Instead, they verify at the end if multiple transactions can complete without affecting each other. They are called validation or certification techniques due to the validation step involved. The typical phases of an OCC method include the read phase (transaction reads the database), validation phase (checks if other transactions changed the read data), and write phase (the transaction can modify the database).

Step-by-step solution

Definition and Differentiation

Optimistic Concurrency Control (OCC) methods assume that multiple transactions can complete without affecting each other, and checks whether this assumption holds true at the end of the transaction. This is different from other techniques like Pessimistic Concurrency Control where locks are used to prevent other users from modifying the data until the transaction is finished.

Understanding the Alternate Names

They are also called validation or certification techniques because they use a validation or certification step in which the system checks if the assumptions, made at the beginning of the transaction, about the system's concurrent state still hold true. If a validation error occurs, the transaction is rolled back and tried again.

Phases of Optimistic Concurrency Control Method

Typical phases of an OCC method include: 1. Read Phase: The transaction reads the database but does not modify it. 2. Validation Phase: The system checks if other transactions have changed the data the current transaction has read. If so, the transaction is rolled back and tried again. 3. Write Phase: If validation is successful, the transaction can modify the database.

Key concepts

Pessimistic Concurrency Control

Pessimistic Concurrency Control (PCC) is a method used to ensure data integrity when multiple transactions occur simultaneously in a database system. This technique assumes that conflicts between concurrent transactions are likely to happen. Therefore, it takes a proactive approach to prevent these conflicts through locking mechanisms.

In PCC, locks are acquired on data items that a transaction needs to access, which prevents other transactions from modifying the same data until the lock is released. This approach minimizes the risk of conflicts and maintains data consistency. However, it can lead to decreased system performance due to increased waiting time when one transaction holds a lock that another needs.

• **Locks**: PCC employs mechanisms like exclusive locks and shared locks to manage access.
• **Blocking**: Transactions may be blocked or delayed, leading to potential bottlenecks.

While PCC is effective in preventing conflicts, it can also be inefficient if transactions are highly interdependent, causing delays. This method benefits systems where transactions are frequent and data consistency is a high priority, such as bank databases or ticket reservation systems.

Validation Techniques

Validation techniques are crucial in optimistic concurrency control methods, used to ensure that transactions do not interfere with each other, after their execution is assumed to be conflict-free.

During the validation phase, the system checks whether any other transactions have made changes to the data that could conflict with the current transaction's view. If no conflicts are detected, the transaction proceeds to commit its changes. However, if conflicts arise, the transaction must be aborted and retried.

• **Conflict Detection**: Looks for overlaps in read-write sets of transactions.
• **Rollback and Retry**: Transactions are aborted and restarted if conflicts are found during validation.

This phase is where optimistic concurrency control methods earn their "validation" or "certification" moniker, as transactions are certified only if they pass the validation step. The advantage here is reduced need for locking, thereby potentially increasing throughput in systems where transaction conflicts are rare.

Transaction Management

Transaction Management is the backbone of database systems, ensuring that all database operations are executed reliably and safely. It involves overseeing multiple operations to ensure that they are processed correctly and that the database remains in a consistent state.

A transaction represents a unit of work that must either complete fully or not at all — a principle known as atomicity. This guarantees that if an error occurs during a transaction, the database will not be left in a partial or incorrect state, as the transaction will be rolled back.

• **Atomicity**: Ensures all or nothing principle in transaction execution.
• **Consistency**: Maintains data integrity according to rules after transaction completion.
• **Isolation**: Transactions are protected from the interference of others until they are finished.
• **Durability**: Completed transactions are persistently saved even if the system fails.

Transaction management protocols, such as the two-phase commit and various concurrency control methods like optimistic and pessimistic controls, handle these properties to provide robust database interaction capabilities. Effective transaction management ensures reliable data handling, which is critical for businesses relying on concurrent data processes.