3 ms·
Made a online tool that tests for this vuln: https://lab.xpaw.me/MS15-034/?host=bing.com https://lab.xpaw.me/MS15-034/?host=bing.com
by xPaw 12y ago
Made a online tool that tests for this vuln: https://lab.xpaw.me/MS15-034/?host=bing.com https://lab.xpaw.me/MS15-034/?host=bing.com
- 51Cards 12y agoTested against three servers, one patched and two not. Small test but it correctly identified which was which.
- el_duderino 12y agoWhat did it say for the ones not vulnerable?
- thecrappycoder 12y ago<sitename> is vulnerable.
- h43k3r 12y agoI was wondering, is their a way to know the the type of webserver a website is using ? I mean can we do it using some kind of headers?
- getsat 12y agoThe "Server" response header will usually tell you but may omit a specific version depending on configuration. Use "curl -I whatever.com" to send a HEAD request and look at the headers in the response.
- e12e 12y agoYou can also try Nmap, but apparently it's not able to tell too much if the server doesn't reply with a header: nmap -T5 -sV --version-all -p 80,443 www.google.com Starting Nmap 6.00 ( http://nmap.org ) at 2015-04-16 03:02 CEST Nmap scan report for www.google.com (80.202.12.244) Host is up (0.0015s latency). Other addresses for www.google.com (not scanned): (...) rDNS record for 80.202.12.244: cache.google.com PORT STATE SERVICE VERSION 80/tcp open http Google httpd 2.0 (GFE) 443/tcp open ssl/http Google httpd 2.0 (GFE) Service Info: OS: Linux; CPE: cpe:/o:linux:kernel Apparently stackoverflow (well know user of .net stack) is "unknown", but microsoft.com gives: nmap -T5 -sV --version-all -p 80,443 microsoft.com Starting Nmap 6.00 ( http://nmap.org ) at 2015-04-16 03:05 CEST Nmap scan report for microsoft.com (134.170.188.221) Host is up (0.18s latency). Other addresses for microsoft.com (not scanned): 134.170.185.46 rDNS record for 134.170.188.221: microsoftproductionstudios.org PORT STATE SERVICE VERSION 80/tcp open http Microsoft IIS httpd 8.5 443/tcp open ssl/http Microsoft IIS httpd 8.5 Service Info: OS: Windows; CPE: cpe:/o:microsoft:windows I didn't look to carefully at the so-output -- maybe there's a funky loadbalancer in front or something.