golang-concurrency

To prevent goroutine leaks, ensure every goroutine has a clear exit strategy using context cancellation, wait groups, or explicit shutdown signals. This guarantees structured concurrency and safe goroutine lifecycles without orphaned processes.

Use channels for passing ownership of data between goroutines, and use mutexes or atomics for protecting shared memory. This Skill guides practical concurrency decisions to ensure safe data passing and avoid unsynchronized memory access.

Implement safe worker pools and pipelines by enforcing sender-owned channel closure, correct channel direction, and bounded concurrency. This prevents common pipeline failures and ensures safe value passing across concurrent stages.

Using time.After in hot loops causes concurrency issues because it allocates a new timer and channel every iteration without being garbage collected until it fires. This creates memory leaks and degrades performance in long-running goroutines.

Handle context cancellation by always including a ctx.Done() case in your select loops. This ensures goroutines terminate promptly when the context is canceled, preventing blocked sends or receives and avoiding goroutine leaks.

errgroup is preferred over WaitGroup when you need bounded concurrency and automatic error propagation. It simplifies managing concurrent goroutines by setting a concurrency limit and canceling the context on the first error.