Hi,

What is a recommended way of hosting or creating a mirror of a website that is blocked in a country. Any suggestions are greatly welcome.

Do you have any restrictions, special needs? The first and straightforward way is to set up a reverse proxy for the website, for example, with nginx.

It is for a Joomla website.

Something like this.

Not sure if it’s what you’re looking for, but there are some (paid) hosting services that are supposed to provide high availability (by deploying mirrors, proxies, etc.). I don’t have personal experience with how well they work.

• Secure Hosting – Qurium Media Foundation
  • e.g. Azadliq.info reloaded – Qurium Media Foundation, Internet blocking in Jordan – Qurium Media Foundation
• https://deflect.ca/

Depending on how much control you have over the server, and how targeted the blocking is, Geneva has a post/paper on server-side evasion.

Thanks for all the suggestions.

I am trying to find an “easy” way to set a mirror of a website.
@ValdikSS If I set a reverse Nginx proxy where shall I point they proxy to? Wouldn’t that proxied IP not get blocked too?

The reverse proxy has to be located outside the censor’s zone of control. That way there is an uncensored path between the reverse proxy and the original web site.

There are some third-party proxies that can sometimes be used to get around simple blocks. (Won’t work with sites requiring login, etc.) For example,

• https://via.hypothes.is/https://example.com/
• https://web.archive.org/web/*/https://example.com
• Google Web Light used to work (link from 2016), but I cannot get it to work today, even putting the browser into mobile device emulation mode.

I was looking into the idea of web mirrors again when I found this topic.
If your mirror doesn’t need to download larges files, you can use Cloudflare workers & pages.
Unfortunately, it seems workers.dev is blocked in Russia, mainland China, Turkmenistan and other countries.

Example: https://community.openguides.workers.dev.

worker.js example

addEventListener('fetch', event => {
  event.respondWith(handleRequest(event.request))
})

async function handleRequest(request) {
  const url = new URL(request.url)
  url.hostname = 'community.torproject.org'  // Set the hostname

  const response = await fetch(url.toString())
  const contentType = response.headers.get('Content-Type')

  // Check if the response is HTML
  if (contentType && contentType.includes('text/html')) {
    let text = await response.text()
    
    // Rewrite URLs in the response body
    text = text.replace(/community\.torproject\.org/g, 'community.openguides.workers.dev')

    return new Response(text, {
      headers: response.headers,
      status: response.status,
      statusText: response.statusText
    })
  } else {
    // Return the response as-is for non-HTML content
    const responseClone = response.clone()
    return new Response(responseClone.body, {
      headers: response.headers,
      status: response.status,
      statusText: response.statusText
    })
  }
}

del
i wrote about google translate here before, but it turns out that it uses google global cache servers in russia (where i am) and they can’t open blocked websites

image745×456 24.7 KB

image736×645 15.7 KB

i remember that this method worked in turkmenistan some time ago, but now the ip address for translating websites is blocked

IPFS allows publishing static websites in a federated (different ipfs gateways are federations) or distributed way (local ipfs-node communicates with the rest of the ipfs network). The last time I looked at ipfs was long ago and unfortunately this approach was easy to be blocked.

The steps:

1. Create a static website from your Joomla target.
2. Distribute static website via ipfs.

See:

1. GitHub - ipfs/distributed-wikipedia-mirror: Putting Wikipedia Snapshots on IPFS
2. GitHub - ballerburg9005/wget-2-zim: creates ZIM files for Kiwix from arbitrary websites with wget and some nifty tricks (doesn't need ServiceWorkers) (zim-files are easily converted to a static website, comment out the code that bundles downloaded static site files into a zim-file).

Freenet https://en.wikipedia.org/wiki/Freenet