goroutine pool

A simple and efficient pool to recycle goroutines instead of creating new ones on the fly. Supports aggregating data into a buffered channel. The caller will wait for all the workers to complete and then get all the responses from every worker into the buffered channel.

Getting Started

Installing

To start using goroutine-pool, install Go and run go get:

$ go get -u github.com/vaibhavmew/goroutine-pool

Usage

Check the following examples to understand the usage submit submit and aggregate

Features

1. Get the response from p.Submit(), once the request is submitted.
2. Aggregate data from all the workers. Pass a buffered channel to the pool. The data will be inserted into it.
3. Supports all data types in request and response. Just modify the request and response structs.
4. Supports timeout. No worker is blocked infinitely.
5. The pool doesn't use queues, sync.Pool or any other algorithm to fetch and return the workers.
6. No use of interface{} or it's conversion anywhere.
7. Instead of passing a function into the pool, we pass the request body.

Increase the pool size at runtime using

p, err := pool.New(5)
if err != nil {
    panic(err)
}

err = p.Tune(10) //here 10 refers to the new pool size
if err != nil {
    panic(err)
}

Close the pool once the workers are done

err = p.Close()
if err != nil {
    panic(err)
}

Checking the error returned by a worker

response := p.Submit(request)
if response.Err != nil {
    fmt.Println(response.Err)
}