Docs

How the proxy works.

For your static IPs.

Your app sends the calls that need a fixed IP through an HTTPS proxy at your pair. The proxy checks the destination against your allowlist and connects; your partner sees your static IP.

One request, step by step.

  1. Your app opens a TLS connection to the proxy on port 3129 at your pair’s IP, and sends your username and password inside it.
  2. It asks the proxy for a connection to the destination with HTTP CONNECT, for example api.partner.example:443.
  3. The proxy checks the host against your pair’s allowlist: exact host names, port 443. Anything else is refused.
  4. Your app starts its TLS session with the destination over that connection. The destination sees the request come from your static IP.
Sequence

HTTP CONNECT through the proxy · Example

# 1. Your app opens TLS to the proxy on port 3129 and logs in (USER:PASS)# 2. Inside it, your app asks for the destination:CONNECT api.partner.example:443 HTTP/1.1Host: api.partner.example:443Proxy-Authorization: Basic <USER:PASS, base64> # 3. The proxy checks the host against your allowlist and answers:HTTP/1.1 200 Connection established # 4. Your app starts its TLS session with api.partner.example over that connection.#    api.partner.example sees the request come from your static IP.

Only HTTPS goes through the proxy, and only as CONNECT. Calls you don’t send through the proxy keep leaving from your platform’s usual addresses.

Which IP a request leaves from.

Two dedicated static IPs on separate servers in separate AWS availability zones. Each IP has its own proxy endpoint: a request through endpoint A leaves from IP A, and through endpoint B from IP B.

Allowlist both IPs with your partner. If a connection to one endpoint fails, your app retries through the other: it does this itself, as in the example under Environment variables.