Pattern Design Concurrency Patterns 2

Question 1

The Half-Sync/Half-Async pattern separates:

Accepted Answer

Synchronous and asynchronous processing into distinct architectural layers

Answer

Half-Sync/Half-Async divides processing into an asynchronous layer (for I/O events) and a synchronous layer (for business logic), connected by a queue.

Question 2

Which concurrency construct is used to block a set of threads until all of them have reached a common synchronization point before any can continue?

Accepted Answer

Barrier

Answer

A Barrier (or CountDownLatch) holds all participating threads at a meeting point until the last thread arrives, then releases them all simultaneously.

Question 3

The Reactor pattern is most commonly used for:

Accepted Answer

Event-driven I/O handling that multiplexes and dispatches service requests from concurrent connections

Answer

The Reactor pattern demultiplexes I/O events from multiple clients and dispatches them synchronously to registered handlers, enabling scalable single-threaded concurrency.

Question 4

In the Thread-Specific Storage (Thread-Local Storage) pattern, each thread has:

Accepted Answer

Its own independent copy of a variable, invisible to other threads

Answer

Thread-Specific Storage gives each thread its own private copy of a variable so threads never contend over or interfere with each other's data.

Question 5

Which concurrency pattern is specifically designed to ensure a lazily initialized shared resource is created exactly once in a multithreaded environment while minimizing locking overhead?

Accepted Answer

Double-Checked Locking

Answer

Double-Checked Locking avoids acquiring a lock on every access by first checking whether the resource is already initialized before entering the synchronized block.

Pattern Design Practice Test

Pattern Design Practice Test

Pattern Design Concurrency Patterns 2