One thousands (1000) HTTP requests: async using NodeJs vs sync using Python vs using Python (with threads)

send.js

// npm install request

let request = require('request')
let range = n => Array.from(Array(n).keys())

data = range(1000)

data.forEach(function (item) {
  request.get("https://httpbin.org/ip", function (error, response, body){
      console.log("Request " + item + " complete.")
  });
})

send.py

# pip install requests

import requests

data = range(1000)

for item in data:
    requests.get("https://httpbin.org/ip")
    print("Request {} complete.".format(item))

send_thread.py

# pip install requests

import requests
from concurrent import futures

data = range(1000)


def send_request(item):
    requests.get("https://httpbin.org/ip")
    print("Request {} complete.".format(item))


executor = futures.ThreadPoolExecutor(max_workers=25)

futures_ = [executor.submit(send_request, x) for x in data]

for f in futures.as_completed(futures_):
    f.result()

time.md

Results

TL;DR Async is very fast 🚀

But with threading, Python gives similar speed.

Async

$ time node send.js

real  0m15.233s
user  0m3.500s
sys  0m0.583s

Sync

$ time python send.py

real  11m38.011s
user  0m19.493s
sys  0m1.433s

Threads

$ time python send_thread.py

real  0m23.800s
user  0m25.393s
sys	 0m3.243s