stream.js

const fs    = require('fs')

const video = './videos/sample.mp4' // source video
let stat = fs.statSync(video)
let range = request.headers.range // will be undefined on first request

if (range) { // if the user is requesting to scrub to a point in the video
  let [start, end] = range.replace(/bytes=/, '').split('-') // the incoming range typically looks like bytes=3214-5435, so extract the numbers out from the request
  start = parseInt(start, 10) // check we are dealing with an integer
  end = end ? parseInt(end, 10) : stat.size - 1 // a request can be for a segment, or could be 'bytes=324-' idicating from here, until the end of the video. Therefore end can be undefined. 
  // if it is undefined set it to the length of the video content
  response.writeHead(206, { // 206 tells the browser to expect a partial request, so a continuation of a stream
    "Content-Range": `bytes ${start}-${end}/${stat.size}`, // the current requested segment
    "Accept-Ranges": "bytes",
    "Content-Length": (end - start) + 1, // not video size, but instead the actual length of buffer
    "Content-Type": "video/mp4"
  }) 

  fs.createReadStream(video, { start, end }).pipe(response) // only stream part of video from given timestamps - handles request for us
} else { // a new first time request would hit this point, since there is no range to scrub to. Therefore just begin a new stream
  response.writeHead(200, {
    "Content-Length": stat.size, // the total length of the content
    "Content-Type": "video/mp4"
  })

  fs.createReadStream(video).pipe(response)
}

It's very awsome