A Python library to utilize AWS API Gateway's large IP pool as a proxy to generate pseudo-infinite IPs for web scraping and brute forcing.
This library will allow the user to bypass IP-based rate-limits for sites and services.
X-Forwarded-For headers are automatically randomised and applied unless given. This is because otherwise, AWS will send the client's true IP address in this header.
AWS' ApiGateway sends its requests from any available IP - and since the AWS infrastructure is so large, it is almost guarenteed to be different each time. By using ApiGateway as a proxy, we can take advantage of this to send requests from different IPs each time. Please note that these requests can be easily identified and blocked, since they are sent with unique AWS headers (i.e. "X-Amzn-Trace-Id").
This package is on pypi so you can install via any of the following:
pip3 install requests-ip-rotatorpython3 -m pip install requests-ip-rotator
import requests
from requests_ip_rotator import ApiGateway
# Create gateway object and initialise in AWS
gateway = ApiGateway("https://site.com")
gateway.start()
# Assign gateway to session
session = requests.Session()
session.mount("https://site.com", gateway)
# Send request (IP will be randomised)
response = session.get("https://site.com/index.html")
print(response.status_code)
# Delete gateways
gateway.shutdown()import requests
from requests_ip_rotator import ApiGateway
with ApiGateway("https://site.com") as g:
session = requests.Session()
session.mount("https://site.com", g)
response = session.get("https://site.com/index.html")
print(response.status_code)Please remember that if gateways are not shutdown via the shutdown() method when using method #1, you may be charged in future.
API Gateway is free for the first million requests per region, which means that for most use cases this should be completely free.
At the time of writing, AWS charges ~$3 per million requests after the free tier has been exceeded.
If your requests involve data stream, AWS would charge data transfer fee at $0.09 per GB.
It is recommended to setup authentication via environment variables. With awscli, you can run aws configure to do this, or alternatively, you can simply set the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY variables yourself.
You can find your access key ID and secret by following the official AWS tutorial.
The ApiGateway class can be created with the following optional parameters:
| Name | Description | Required | Default |
|---|---|---|---|
| site | The site (without path) requests will be sent to. | True | |
| regions | An array of AWS regions to setup gateways in. | False | ip_rotator.DEFAULT_REGIONS |
| access_key_id | AWS Access Key ID (will override env variables). | False | Relies on env variables. |
| access_key_secret | AWS Access Key Secret (will override env variables). | False | Relies on env variables. |
from ip_rotator import ApiGateway, EXTRA_REGIONS, ALL_REGIONS
# Gateway to outbound HTTP IP and port for only two regions
gateway_1 = ApiGateway("http://1.1.1.1:8080", regions=["eu-west-1", "eu-west-2"])
# Gateway to HTTPS google for the extra regions pack, with specified access key pair
gateway_2 = ApiGateway("https://www.google.com", regions=EXTRA_REGIONS, access_key_id="ID", access_key_secret="SECRET")
An ApiGateway object must then be started using the start method.
By default, if an ApiGateway already exists for the site, it will use the existing endpoint instead of creating a new one.
This does not require any parameters, but accepts the following:
| Name | Description | Required |
|---|---|---|
| force | Create a new set of endpoints, even if some already exist. | False |
| endpoints | Array of pre-existing endpoints (i.e. from previous session). | False |
# Starts new ApiGateway instances for site, or locates existing endpoints if they already exist.
gateway_1.start()
# Starts new ApiGateway instances even if some already exist.
gateway_2.start(force=True)
Requests are sent by attaching the ApiGateway object to a requests Session object.
The site given in mount must match the site passed in the ApiGateway constructor.
import requests
# Posts a request to the site created in gateway_1. Will be sent from a random IP.
session_1 = requests.Session()
session_1.mount("http://1.1.1.1:8080", gateway_1)
session_1.post("http://1.1.1.1:8080/update.php", headers={"Hello": "World"})
# Send 127.0.0.1 as X-Forwarded-For header in outbound request (otherwise X-Forwarded-For is randomised).
session_1.post("http://1.1.1.1:8080/update.php", headers={"X-Forwarded-For", "127.0.0.1"})
# Execute Google search query from random IP
session_2 = requests.Session()
session_2.mount("https://www.google.com", gateway_2)
session_2.get("https://www.google.com/search?q=test")
It's important to shutdown the ApiGateway resources once you have finished with them, to prevent dangling public endpoints that can cause excess charges to your account.
This is done through the shutdown method of the ApiGateway object. It will close all resources for the regions specified in the ApiGateway object constructor.
# This will shutdown all gateway proxies for "http://1.1.1.1:8080" in "eu-west-1" & "eu-west-2"
gateway_1.shutdown()
# This will shutdown all gatewy proxies for "https://www.google.com" for all regions in ip_rotator.EXTRA_REGIONS
gateway_2.shutdown()Alternatively, you can selectively shutdown specific endpoints, if needed. To do this, simply pass in an array of endpoints to the shutdown() method, i.e:
# This will force start a new gateway (i.e. create new endpoints even if some exist on the region already), and then delete the first 3 of them only.
gateway_3 = ApiGateway("http://1.1.1.1:8082", regions=ALL_REGIONS)
endpoints = gateway_3.start(force=True)
gateway_3.shutdown(endpoints[:3])The core gateway creation and organisation code was adapter from RhinoSecurityLabs' IPRotate Burp Extension.
The X-My-X-Forwarded-For header forwarding concept was originally conceptualised by ustayready in his fireprox proxy.