Waiting in C# is a core skill for writing responsive and reliable applications. Instead of busy loops that burn CPU, you use asynchronous coordination tools to pause work until a condition is met.
This guide walks through practical patterns with clear examples, performance notes, and common pitfalls. You will learn how to coordinate threads and tasks while keeping code robust and maintainable.
| Topic | Key Constructs | Use Case | Best Practice |
|---|---|---|---|
| Thread-level waiting | Monitor.Wait, Pulse, Mutex | Classic producer-consumer on shared memory | Always wait inside a lock on the same object |
| Task-level waiting | await, Task.Delay, Task.WhenAny | Nonblocking I/O, timeouts, cancellation | Avoid Task.Result; use await to keep scalability |
| Async coordination | SemaphoreSlim, AsyncManualResetEvent | Rate limiting, signaling across async methods | Prefer async-compatible primitives over blocking ones |
Thread Synchronization Basics
Locking and Monitor Patterns
Thread waiting often starts with locking to protect shared state. Inside a lock, you can use Monitor.Wait to yield the lock and sleep until another thread calls Monitor.Pulse.
Avoiding Race Conditions
Always check a condition in a loop after waking up, because spurious wakeups can occur. This ensures you only proceed when the expected state is truly ready.
Task and Async Coordination
Using Task.Delay for Timeouts
Task.Delay returns a task that completes after a time span. When combined with Task.WhenAny, you can implement timeouts without blocking threads.
CancellationToken for Cooperative Cancellation
Pass a CancellationToken to long-running async work so that waiting tasks can be canceled cleanly instead of being abandoned mid-wait.
Async-Friendly Synchronization Primitives
SemaphoreSlim for Controlled Concurrency
SemaphoreSlim lets you throttle concurrent operations while supporting both synchronous and asynchronous waits efficiently.
Channel and Dataflow Waiting
Channels offer built-in waiting semantics for reading and writing, which simplify producer-consumer scenarios without manual locking.
Key Takeaways
- Use Monitor.Wait and Pulse only inside lock blocks to safely yield and resume threads.
- Prefer async-compatible primitives like SemaphoreSlim and Channels for scalable waiting.
- Always apply timeouts via Task.WhenAny and CancellationToken to keep systems responsive.
- Avoid blocking on asynchronous tasks to prevent deadlocks and thread pool starvation.
- Validate state in a loop after wakeups to handle spurious signals and avoid race conditions.
FAQ
Reader questions
What happens if I call Monitor.Wait without a Pulse in C#?
The thread will wait indefinitely, so ensure every Wait has a matching Pulse on the same lock object to eventually resume progress.
How can I wait with a timeout using async code in C#?
Use Task.WhenAny with Task.Delay and a cancellation token. If the delay task finishes first, you know the operation timed out.
Can I block a thread safely while waiting for an async operation in C#?
Avoid blocking on async code. If you must, use a dedicated thread or configure await with false to prevent deadlocks, but prefer fully async designs.
What is the difference between Task.Delay and Thread.Sleep for waiting in C#?
Task.Delay is nonblocking and ideal for async workflows, while Thread.Sleep blocks the current thread and should be avoided in service code.