Popularity
0.5
-
Activity
4.2
-
3
1
0
Programming language: Go
Tags:
Distributed Systems
consistenthash alternatives and similar packages
Based on the "Distributed Systems" category
-
tendermint
High-performance middleware for transforming a state machine written in any programming language into a Byzantine Fault Tolerant replicated state machine using the Tendermint consensus and blockchain protocols. -
glow
Easy-to-Use scalable distributed big data processing, Map-Reduce, DAG execution, all in pure Go. -
emitter-io
High performance, distributed, secure and low latency publish-subscribe platform built with MQTT, Websockets and love. -
gleam
Fast and scalable distributed map/reduce system written in pure Go and Luajit, combining Go's high concurrency with Luajit's high performance, runs standalone or distributed. -
Olric
Distributed cache and key/value store. It can be used both as an embedded Go library and as a language-independent service. -
go-sundheit
A library built to provide support for defining async service health checks for golang services. -
celeriac
A library for adding support for interacting and monitoring Celery workers, tasks and events in Go
Do you think we are missing an alternative of consistenthash or a related project?
Popular Comparisons
|
SaaSHub - Software Alternatives and Reviews
View more from us
|
README
Consistent Hashing
A Go library that implements Consistent Hashing
This package is implemented based on golang/groupcache package with some improvements
Definitions in this README:
node: Refers to the key which is going to be stored in the hash ring or hash table
rKey: Refers to the request hash which is not going to be stored, it's used to find the upper closest node in the hash ring to the rKey
Improvements:
Removefunction added - sort and remove a node from the hash ring- int hashes replaced with uint32
- Number of replicas is now configurable while adding new node
Note: This is useful when capacity is not the same for all nodes
Usage
import (
chash "github.com/mbrostami/consistenthash"
)
// Create ConsistentHash with 2 replicas
ch := chash.NewConsistentHash(2, nil)
ch.Add("127.0.0.1:1001") // node 1
ch.Add("127.0.0.1:1002") // node 2
ch.AddReplicas("127.0.0.1:1003", 4) // node3 has more capacity so possibility to get assigned request is higher than other nodes
rKey := "something like request url"
node := ch.Get(rKey) // find upper closest node
fmt.println(node) // this will print out one of the nodes
Test
Benchmark
go test -bench=.
Race Condition
go test --race
Thanks to Mohammad Rajabloo for reporting the race issue
Tests
go test