mivinci/mc

A simple distributed caching database for learning purposes.

MC

MC - short for Memory Cache - is a simple distributed caching server for learning purpose.

Features

• In memory LRU cache
• RPC transport
• Cluster
• Command line client
• Go client

So far, MC has only get, add and remove of caching and a Consistent Hash Algorithm will be considered to use for the scalability in the future.

Usage

Install

go get github.com/mivinci/mc

Or (not available yet)

docker pull mivinci/mc

Or download the pre-build version for macOS or Windows.

Run

mc can be run either a server or a client.

Run an MC server.

mc -server

By default, MC chooses 127.0.0.1:8000 to serve on, or you can decide the address by providing a -address parameter.

mc -server -address 127.0.0.1:8001

for short

mc -server -address :8001

You can customize the maximum caching size for a single MC server.

mc -server -c 128

Run an MC client.

mc -client

Also, the MC client chooses 127.0.0.1:8000 to dial during usage. You can provide an -address parameter to the MC client to communicate with a different address.

Use Cluster

Once you have run multiple MC servers, you can communicate with them simply by using the MC client with multiple addresses seperated by commas.

mc -client -address 1.1.1.1:8000,1.1.1.2:8001,1.1.1.3:8002

The MC command line client will lead you to an interface like this.

$ mc -client
MC v1.0.0, press [ctrl c] to exit.
127.0.0.1:8000> get name
XJJ
127.0.0.1:8000> remove name
127.0.0.1:8000> get name
nil

Go Client

import "github.com/mivinci/mc"

Add

r := mc.Add("name", []byte("XJJ"))

Get

r := mc.Get("name")

Remove

r := mc.Remove("name")

Cluster

client := mc.NewClient("1.1.1.1:8000", "1.1.1.2:8001", "1.1.1.3:8000")
r := client.Get("name")
...

About

Thanks to the projects and videos below for the inspiration.

• https://github.com/asim/mq

• https://www.youtube.com/watch?v=iuqZvajTOyA