What is sync.Pool and when should you use it?

sync.Pool is a per-processor cache of reusable objects that reduces GC pressure by allowing frequently allocated short-lived objects to be reused rather than garbage collected.

sync.Pool for buffer reuse

sync.Pool guidelines

1
Objects CAN be evicted at any GC cycle — never rely on persistence across requests
2
Always Reset/Clear objects before returning them to the pool — stale state causes bugs
3
Best for: byte buffers, encoder/decoder instances, temporary structs in hot paths
4
The standard library uses Pool internally: fmt package reuses printer objects, json uses encoder pools
5
Measure impact with benchmarks: sync.Pool is only beneficial when allocation cost is proven high