workerpool alternatives and similar packages
Based on the "Goroutines" category.
Alternatively, view workerpool alternatives based on common mentions on social networks and blogs.
-
semaphore
Semaphore pattern implementation with timeout of lock/unlock operations based on channel and context. -
neilotoole/errgroup
Drop-in alternative to sync/errgroup, limited to a pool of N worker goroutines. -
Hunch
Hunch provides functions like: All, First, Retry, Waterfall etc., that makes asynchronous flow control more intuitive. -
goccm
Go Concurrency Manager package limits the number of goroutines that allowed to run concurrently. -
conexec
A concurrent toolkit to help execute funcs concurrently in an efficient and safe way.It supports specifying the overall timeout to avoid blocking and uses goroutine pool to improve efficiency. -
go-tools/multithreading
Manage a pool of goroutines using this lightweight library with a simple API. -
queue
Gives you a sync.WaitGroup like queue group accessibility. Helps you to throttle and limit goroutines, wait for the end of the all goroutines and much more. -
hands
A process controller used to control the execution and return strategies of multiple goroutines. -
concurrency-limiter
Concurrency limiter with support for timeouts , dynamic priority and context cancellation of goroutines. -
oversight
Oversight is a complete implementation of the Erlang supervision trees.
Get performance insights in less than 4 minutes
Do you think we are missing an alternative of workerpool or a related project?
Popular Comparisons
-
workerpoolvstunny
-
workerpoolvssemaphore go
-
workerpoolvsgoworker
-
workerpoolvssemaphore
-
workerpoolvssworker
|
SaaSHub - Software Alternatives and Reviews
Sponsored
www.saashub.com
|
README
π¨βπ§ worker-pool
Go simple async worker pool.
π ABOUT
Worker pool is a software design pattern for achieving concurrency of task execution. Maintains multiple workers waiting for tasks to be allocated for concurrent execution. By maintaining a pool of workers, the model increases performance and avoids latency in execution. The number of available workers might be tuned to the computing resources available.
You can read more about worker pools in Go here.
Contributors:
Want to contribute ? Feel free to send pull requests!
Have problems, bugs, feature ideas? We are using the github issue tracker to manage them.
π Documentation
For examples visit godoc#pkg-examples
For GoDoc reference, visit pkg.go.dev
π HOW TO USE
π Benchmark
CPU: 3,3 GHz Intel Core i7
RAM: 16 GB 2133 MHz LPDDR3
β worker-pool git:(master) β go test -bench=. -cpu=4 -benchmem
goos: darwin
goarch: amd64
pkg: github.com/vardius/worker-pool/v2
BenchmarkWorker1-4 3944299 284 ns/op 56 B/op 3 allocs/op
BenchmarkWorker1Parallel-4 7394715 138 ns/op 48 B/op 2 allocs/op
BenchmarkWorker100-4 1657569 693 ns/op 56 B/op 3 allocs/op
BenchmarkWorker100Parallel-4 3673483 368 ns/op 48 B/op 2 allocs/op
BenchmarkWorkerNumCPU-4 2590293 445 ns/op 56 B/op 3 allocs/op
BenchmarkWorkerNumCPUParallel-4 3591553 298 ns/op 48 B/op 2 allocs/op
PASS
ok github.com/vardius/worker-pool/v2 9.511s
π« Basic example
package main
import (
"fmt"
"sync"
"github.com/vardius/worker-pool/v2"
)
func main() {
var wg sync.WaitGroup
poolSize := 1
jobsAmount := 3
workersAmount := 2
// create new pool
pool := workerpool.New(poolSize)
out := make(chan int, jobsAmount)
worker := func(i int) {
defer wg.Done()
out <- i
}
for i := 1; i <= workersAmount; i++ {
if err := pool.AddWorker(worker); err != nil {
panic(err)
}
}
wg.Add(jobsAmount)
for i := 0; i < jobsAmount; i++ {
if err := pool.Delegate(i); err != nil {
panic(err)
}
}
go func() {
// stop all workers after jobs are done
wg.Wait()
close(out)
pool.Stop() // stop removes all workers from pool, to resume work add them again
}()
sum := 0
for n := range out {
sum += n
}
fmt.Println(sum)
// Output:
// 3
}
π [License](LICENSE.md)
This package is released under the MIT license. See the complete license in the package
*Note that all licence references and agreements mentioned in the workerpool README section above
are relevant to that project's source code only.