https://github.com/node-m2m/edge-load-balancing
A quick demo on how to create a load balancer server for edge computing.
https://github.com/node-m2m/edge-load-balancing
Last synced: 3 months ago
JSON representation
A quick demo on how to create a load balancer server for edge computing.
- Host: GitHub
- URL: https://github.com/node-m2m/edge-load-balancing
- Owner: node-m2m
- License: mit
- Created: 2023-03-21T18:34:18.000Z (about 3 years ago)
- Default Branch: main
- Last Pushed: 2025-02-28T00:55:09.000Z (over 1 year ago)
- Last Synced: 2025-02-28T07:58:06.090Z (over 1 year ago)
- Size: 112 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
## Edge Load Balancer
In this example, the client will monitor the data 'test-data' topic from an edge load balancer server.
The load balancer will fetch the data from 3 sources (server1, server2, and server3) one at a time using a simple round robin method.
It will then send the fetched data one at a time to the client or to every connected clients.
If one of the load balancer 3 sources of data goes down due to some unexpected events, it will continue to provide the data from the other sources.
#### Create a project directory for each edge endpoint and install *m2m*.
```js
$ npm install m2m
```
### Server 1
#### Save the code below as *app.js* in your server1 project directory.
```js
const m2m = require('m2m')
let edge = new m2m.Edge({name:'server1'})
// simulated data source
function dataSource(){
return 20 + Math.floor(Math.random() * 10)
}
m2m.connect(() => {
let port = 8144
edge.createServer(port, '127.0.0.1', (server) => {
server.on('error', (error) => {
console.log('error:', error)
})
server.dataSource('test-data', (tcp) => {
tcp.send({server:1, topic:tcp.topic, value:dataSource()})
})
})
})
```
### Server 2
#### Save the code below as *app.js* in your server2 project directory.
```js
const m2m = require('m2m')
let edge = new m2m.Edge({name:'server2'})
// simulated data source
function dataSource(){
return 20 + Math.floor(Math.random() * 10)
}
m2m.connect(() => {
let port = 8145
edge.createServer(port, '127.0.0.1', (server) => {
server.on('error', (error) => {
console.log('error:', error)
})
server.dataSource('test-data', (tcp) => {
tcp.send({server:2, topic:tcp.topic, value:dataSource()})
})
})
})
```
### Server 3
#### Save the code below as *app.js* in your server3 project directory.
```js
const m2m = require('m2m')
let edge = new m2m.Edge({name:'server3'})
// simulated data source
function dataSource(){
return 20 + Math.floor(Math.random() * 10)
}
m2m.connect(() => {
let port = 8146
edge.createServer(port, '127.0.0.1', (server) => {
server.on('error', (error) => {
console.log('error:', error)
})
server.dataSource('test-data', (tcp) => {
tcp.send({server:3, topic:tcp.topic, value:dataSource()})
})
})
})
```
### Load Balancer
#### Save the code below as *app.js* in your load balancer project directory.
```js
const m2m = require('m2m')
let edge = new m2m.Edge({name:'Load Balancer'})
m2m.connect(() => {
let ec1 = new edge.client(8144, '127.0.0.1')
let ec2 = new edge.client(8145, '127.0.0.1')
let ec3 = new edge.client(8146, '127.0.0.1')
let port = 8143, data = null, round = 1
edge.createServer(port, '127.0.0.1', (server) => {
server.on('connection', (count) => {
console.log('connected client', count)
})
server.publish('test-data', async (tcp) => {
if(round === 1){
round = 2
data = await ec1.read(tcp.topic)
}
else if(round === 2){
round = 3
data = await ec2.read(tcp.topic)
}
else if(round === 3){
round = 1
data = await ec3.read(tcp.topic)
}
tcp.send(data)
})
})
})
```
### Client
#### Save the code below as *app.js* in your client project directory.
```js
const m2m = require('m2m')
let edge = new m2m.Edge({name:'client'})
async function main (){
await m2m.connect();
let ec = new edge.client(8143, '127.0.0.1')
ec.on('ready', (result) => {
console.log('ready:', result)
})
ec.on('error', (error) => {
console.log('error:', error)
})
ec.subscribe('test-data', (data) => {
console.log(data)
})
}
```
#### Start the application on each endpoint.
```js
$ node app.js
```
You should get a client output similar to the result as shown below.
```js
ready: true
{ server: 1, topic: 'test-data', value: 27 }
{ server: 2, topic: 'test-data', value: 25 }
{ server: 3, topic: 'test-data', value: 22 }
...
```