Score:0

Socks5 proxy requests (via Selenium) are nerfed; same requests via cURL works

mk flag

Good Day,

I've seen a lot of questions like this but I haven't seen anything similar to my situation. I am not sure if this is the right or best place to ask.

The issue I am facing is that I have a Python script, using Selenium, to make a request to a website. For everything I have tried I can't find out where the issue lies.

Here's the specific situation:

I have two Digital Ocean droplets running in the same region. DO1 contains my python script, using selenium to .get(URL) a resource.

DO2 contains my socks5 proxy server. Set up with ssh -f -N -D 0.0.0.0:1080 localhost.

Now if I run the following:

  1. DO1 requests https://www.google.com with no proxy = No issues, ~0.8 sec request
  2. DO1 requests https://www.google.com with proxy = No issue, ~1.1 sec request

Now here's the issue.

When I use DO1 to request another website (https://mywebsite2.ru/) through the proxy, it takes ~3 minutes to respond. It does respond, it's super slow, like it's nerfed. And it's a brand new droplet.

The proxy will output, 3 times for every request channel X: open failed: connect failed: Connection timed out

I've tried created a proxy on a new droplet in different region - same effect.

If I try and use DO1 to request directly, same issue. It takes ~3 minutes to respond.

Now I figured the IPs themselves might be tainted. If I use DO1 to curl the same website, I get the results right away. If I use DO1 to curl the website through the proxy, I get the results right away.

Selenium does have a browser header attached to the request, I don't believe curl does by default. I've also tried changing those around.

So I am not sure how to address this. (1) I am not sure what might be causing the nerfing in the python scripts and (2) I am not sure what I can try and explore at this time to find the issue.

Hoping I can be pointed in the right direction.

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.