IS411P Distributed Database Systems Quiz 2

Concurrency Control in Distributed Database Systems

• Concurrency control algorithms enforce the property of serializability in a distributed database system.

Reliability Measures

• The main purpose of reliability measures is to ensure that the database system remains operational even in the presence of failures.

Consistency in Distributed Database Systems

• Consistency primarily refers to the requirement that the database remains in a consistent state, even in the presence of concurrent transactions.

Distributed Concurrency Control

• The main role of distributed concurrency control is to coordinate the access of multiple transactions to shared data items in a distributed database system.

Integrity Enforcement

• Integrity enforcement handles the aspect of ensuring that the database remains in a valid state, by enforcing constraints and rules.

Transactions in Distributed Database Systems

• The primary objective of a transaction is to execute a sequence of operations as a single, all-or-nothing unit of work.

Timestamp-Based Concurrency Control Algorithms

• The main purpose of assigning a unique timestamp to each transaction is to determine the order in which transactions are executed.

Assigning Timestamps in Distributed DBMS

• It is preferable for each site to autonomously assign timestamps based on its local counter to avoid synchronization overhead and to improve performance.

Basic TO Algorithm

• The condition for executing conflicting operations Oij and Okl is that the timestamp of Oij is less than the timestamp of Okl.
• The primary drawback of the basic TO algorithm is that it can lead to cascading aborts.

Conservative TO Algorithms

• The goal of conservative TO algorithms is to reduce the number of aborts and improve system performance.

Assigning Timestamps in Basic TO Algorithm

• The purpose of assigning two timestamps, read timestamp [rts(x)] and write timestamp [wts(x)], to each data item is to ensure that the system can recover from failures.

Deadlocks in Basic TO Algorithm

• The basic TO algorithm never causes deadlocks because it ensures that transactions are executed in a serialized manner.

Assigning Timestamps in Distributed DBMS

• Using a global (system-wide) monotonically increasing counter for assigning timestamps can lead to synchronization overhead and performance issues.
• Each site appends its own identifier to the counter value when autonomously assigning timestamps based on its local counter to avoid confusion and ensure uniqueness.

Using System Clock Values for Timestamps

• The significance of using system clock values instead of counter values for assigning timestamps is that it can provide a more accurate and reliable timestamp.

Coordinating TM in Basic TO Algorithm

• The coordinating TM is responsible for ensuring that all participating sites agree on the outcome of a transaction, and for coordinating the commitment or abort of a transaction.