On Friday, Xavier talked about the newly introduced HTTP Query method. This new method was introduced to allow "GET" requests that include a body. The main reason for this was that GET requests typically do not contain a body. But what if they do?

I did a quick check of a couple of common web servers I had handy, to see what would happen:

Apache

For this test, I ran Apache 2.4.68 on a Mac. It happily accepted a body with a GET request:

% nc -c localhost 8080
GET /cgi-bin/test-cgi HTTP/1.1
Host: localhost
Content-Length: 6

TEST
HTTP/1.1 200 OK
Date: Tue, 22 Sep 2026 14:39:17 GMT
Server: Apache/2.4.68 (Unix)
Transfer-Encoding: chunked
Content-Type: text/plain; charset=iso-8859-1

18a
CGI/1.0 test script report:
[some details omited]
CONTENT_LENGTH = 6
BODY = TEST

The data was collected using a slightly modified version of the standard "test-cgi" script. The body was received just fine, and a 200 status was returned.

NGINX

% nc -c 10.128.1.11 80
GET /cgi-bin/test-cgi HTTP/1.1
Host: localhost
Content-Length: 6

TESTHTTP/1.1 301 Moved Permanently
Server: nginx

The request still did not trigger an error. But the body was ignored. The server started sending the response as soon as it received the headers. The body was ignored.

Python

A simple Python web server (python -m http.server 8000) appears to behave just like NGINX. The body is ignored, but a response is sent back, and the status code is 200. 

Node

Node also ignores the Content-Length header and processes the request without error.

lighthttpd

lighttpd/1.4.74 will return a 400 error and refuse to process the request.

Java/Tomcat

Tomcat ignores the Content-Length header but returns a 200 response.

Do you have any web servers to test to see how they respond to a GET request with a body?

--
Johannes B. Ullrich, Ph.D. , Dean of Research, SANS.edu
Twitter|