Providing a Custom Executor to runAsync
What you'll learn: How to supply your own Executor to runAsync to control which thread pool executes your asynchronous tasks.
The Need for Control
By default, runAsync uses the ForkJoinPool.commonPool() to execute your fire-and-forget tasks. But just like with supplyAsync, you often need more control—whether to isolate workloads, manage thread priority, limit concurrency, or separate I/O-bound from CPU-bound tasks.
The Two-Argument Form
CompletableFuture.runAsync() has an overloaded version that accepts an Executor:
CompletableFuture<Void> future = CompletableFuture.runAsync(
() -> {
System.out.println("Running on: " +
Thread.currentThread().getName());
},
myCustomExecutor // Your ExecutorService here
);
The second parameter is your custom Executor (typically an ExecutorService created via factory methods like newFixedThreadPool or newCachedThreadPool).
Practical Example
ExecutorService ioPool = Executors.newFixedThreadPool(10);
CompletableFuture<Void> future = CompletableFuture.runAsync(() -> {
// Perform I/O operation like writing logs
System.out.println("Logging on dedicated pool");
}, ioPool);
// Don't forget to shut down your custom executor when done
ioPool.shutdown();
Why This Matters
Using a custom executor with runAsync lets you:
- Separate concerns: Keep background logging tasks isolated from critical business logic
- Control resource allocation: Dedicate a specific number of threads to side-effect operations
- Prevent starvation: Avoid blocking the common pool with long-running I/O tasks
Key Takeaway: Supplying a custom Executor to runAsync gives you full control over execution context, letting you optimize thread usage and isolate different workload types in your asynchronous pipeline.