CrabPool | Yet Another Private Network Solution

CrabPool aims to create a centralized network that one can access resources related to any other nodes in the network easily.

Index

• Installation
• Usage
• API
• Misc
• Examples
• Changelog

Installation

npm i crabpool

Usage

To create a CrabPool network, a unique Master is required, and all other Nodes would connect to the Master to accomplish the network.

Create a Master

const CrabPool = require('crabpool')

CrabPool(
{
	Cipher : () => <A cipher, will be described in section API>,
	Deipher : () => <A decipher, responses to the cipher above>,
	PortMaster : 3389 // Port to deploy the master
})

Create a Node

const CrabPool = require('crabpool')
const Net = require('net')

CrabPool(
{
	Cipher : () => <The same cipher used by the master>,
	Deipher : () => <The same decipher used by the master>,
	Pipe : () => Net.createConnection(3389,'IP to the master')
})

Custom the Links

To access the resources of other Nodes, one needs to configure links. Simply install the optional dependence ws by npm i ws, and add PortWeb property to the setup object like this:

const CrabPool = require('crabpool')

CrabPool(
{
	Cipher : ...,
	Deipher : ...,
	Pipe : ...,
	PortWeb : 4000
})

Then a local web server localhost:4000 will be deployed and can be visited from browsers (that support WebSocket). The initial manage token can be found in the Data Folder You can also use it as an Express.Router and mount it to any path you like on an existing server, which will also be described in the API section.

API

CrabPool(Option)

• Option : Object
  • Cipher : import('crypto').Cipher Optional. A cipher to encrypt/compress/etc the data stream, must not use block ciphers that output only when blocks are filled (such as AES-CBC). Leave empty if the socket is already secured or for debug use. See Example : Cipher
  • Decipher : import('crypto').Cipher Optional. A decipher responses to the Cipher above. See Example : Cipher
  • PortMaster : number Required for the Master, must not be present for Nodes. Port to deploy a Master server.
  • PortWeb : number Optional. Port to deploy the Web Server.
  • Pipe(Log,Ind?) Require for Nodes, must not be present for the Master. See Example : Pipe
    • Log : (...Q : any[]) => any. Write logs.
    • Ind? : boolean. True will be passed if this pipe connection is independent for a single socket.
    • Returns : import('net').Socket | WishNS.Provider<import('net').Socket>
  • PipeRetry : number Optional. Default 1E4ms. Interval to wait before connecting to the Master again after the previous connection closed.
  • Tick : number Optional. Default 6E4ms. Heartbeat interval.
  • Timeout : number Optional. Default 2 * Tickms. Time to wait to shutdown an inactive connection.
  • Data : string Optional. Path to store settings, logs and other files. See Data Folder Structure
  • Log : (...Q : any[]) => any Optional. Control how should CrabPool logs.
  • RecordByte : number Optional. Default 0. Record first few bytes of the communications of each link.
• Returns : Object
  • Exp : (Express? : import('express').Router) => import('express').Router. Given an optional Express.Router object, returns the Router. See Example : Custom Web Server
  • Soc : (S : import('ws'),H : import('http').IncomingMessage) => any. Used to handle event import('ws')::on('connection'). See Example : Custom Web Server

Misc

Data Folder Structure

Windows %AppData%/ZED/CrabPool
Unix %HOME%/.local/share/ZED/CrabPool
Mac %HOME%/Library/Preferences/ZED/CrabPool

ZED/CrabPool
+-- ID # Unique machine ID, randomly generated on the first run.
+-- Key # The manage token for the Web Server, randomly generated on the first run
+-- Pool.db # The database
+-- Session.json # Stores the previous connection status
+-- Log/ # All logs will be output here.
|   +-- Event*.log

Examples

Example : Cipher

const Crypto = require('crypto')
const Key = ...
const IV = ...

CrabPool(
{
	Cipher : () => Crypto.createCipheriv('AES-128-CFB',Key,IV),
	Decipher : () => Crypto.createDecipheriv('AES-128-CFB',Key,IV),
})

Example : Pipe

Simply connect to the Master

const CrabPool = require('crabpool')
const Net = require('net')

CrabPool(
{
	Pipe : () => Net.createConnection({host : ...,port : ...})
})

Connect to the Master with dynamic address

const CrabPool = require('crabpool')
const Net = require('net')
const Wish = require('@zed.cwt/wish')

let Addr

CrabPool(
{
	Pipe : Wish.X.WrapPromise((Log,Ind) =>
	{
		if (Ind && Addr) return Promise.resolve(Net.createConnection(Addr))
		return ResolveAddr()
			.then(O =>
			{
				Log('Resolved',O)
				return Net.createConnection(Addr = O)
			})
	})
})

// Can also use RxJS here
CrabPool(
{
	Pipe : Wish.X.WrapRx(Log =>
	{
		return RxJS.Observable<Net.Socket>
	})
})

Example : Custom Web Server

Required dependencies

npm i ws express

Then

const CrabPool = require('crabpool')
const Express = require('express')
const HTTP = require('http')
const WS = require('ws')

const Pool = CrabPool({...})
const Exp = Express().use('/Fish',Pool.Exp())
const Server = HTTP.createServer(Exp).listen(8000)
new WS.Server({server : Server,path : '/Fish/'})
	.on('connection',Pool.Soc)

And now you can visit from localhost:8000/Fish

Changelog

Updates from v0.x.x

V1 is a rewrite of V0, including the following highlights

• Extensible protocol. The size of message packets is up to 65535 in V0 including random paddings, making it impossible to handle larger messages. V1 now supports arbitrary sizes of messages.
• Socket pause & resume. In an asymmetrical connection network for V0, data may be queued in the memory waiting to be sent, causing congestion issues. It is now handled in V1 by pausing & resuming at appropriate timing.
• Database. V1 now uses SQLite instead of JSON files to persist infomation.
• GlobalLink. It is now possible to configure common links at the Master side, the all nodes would benefit from them.
• Statistic. V1 added few statistic fields to monitor the pool.
• Active connection. It is now possible to monitor/cut active connections through the current node. Because of this, we removed the socket timeout facility, so that connections can stay as long as both the server & the client agree.