singleton_task

README

singleton-task

Description

Due to the nature of microservices, k8s etc., we have multiple pods\instances of the same service, but in some cases, there is a the requirement to run the specific job in a single instance, for example, some parsing jobs with rate limiting, database migrations, etc. Here come the leader election algorithms.

This implementation is targeted for redlock approach

Installation

go get github.com/skynet2/singleton-task

Quickstart

package main

import (
	"context"
	"fmt"
	"os"
	"os/signal"
	"syscall"
	"time"

	"github.com/bsm/redislock"
	"github.com/redis/go-redis/v9"
	singletonTask "github.com/skynet2/singleton-task"
)

func main() {
	sig := make(chan os.Signal, 1)
	signal.Notify(sig, syscall.SIGINT, syscall.SIGTERM)

	instance := singletonTask.NewSingletonRedLock(redislock.New(redis.NewClient(&redis.Options{
		Network: "tcp",
		Addr:    "127.0.0.1:6379",
	})), "long-running-job", func(ctx context.Context) {
		for ctx.Err() == nil {
			fmt.Println("dequeue from queue and send http request.")
			time.Sleep(1 * time.Second)
		}
	}, context.Background(), 30*time.Second)

	if err := instance.StartAsync(); err != nil {
		panic(err)
	}

	fmt.Println("awaiting signal")
	<-sig
	_ = instance.Close()
}

More examples can be found inside tests

Documentation

Index

• type Fn
• type Singleton
  • func NewSingletonRedLock(locker *redislock.Client, key string, fn Fn, ctx context.Context, ...) Singleton

Types

type Fn

type Fn func(ctx context.Context) error

type Singleton

type Singleton interface {
	Close() error
	StartAsync() error
}

func NewSingletonRedLock

func NewSingletonRedLock(
	locker *redislock.Client,
	key string,
	fn Fn,
	ctx context.Context,
	ttl time.Duration,
) Singleton