Awesome
VoIP Patrol
Important note:
This is a fork of the original project with some breaking changes
- Changed format of JSON reports to support VOTLS
- Added possibility to record calls to have the possibility to test media after
- Reworked STUN options to support
DTLS
media more accurate - Extended
expected_XXX
parameters for accept/call tests - Reworked match mechanism on
accept
test to rely on Contact URI parameters (seeaccount
-match_account
notes below) - Added
fail_on_accept
parameter to control calls, that should not happen - Code formatting :)
In general, I'm trying to follow the original project features and changes, but don't expect, that your existing advanced scenarios from the original project will work with this fork, but the simple ones will work without any modifications.
VoIP signaling and media test automaton
Designed to automate end2end and or integration tests.
VoIP patrol will follow a scenario in XML format and will output results in JSON.
Each line in the output file is a separate JSON structure, note that the entire file is not a valid JSON file, this is because VoIP patrol will output results as they become available.
It is possible to test many scenarios that are not easy to test manually like a re-invite with a new codec.
This version is extension of original project and contains changes (in reports and configuration), that are not compatible with original version
Docker quick start
Linux Debian building from sources
Load test example
run
./voip_patrol --help
Example: making a test call
<config>
<actions>
<action type="call" label="us-east-va"
transport="tls"
expected_cause_code="200"
caller="15147371787@noreply.com"
callee="12012665228@target.com"
to_uri="+12012665228@target.com"
max_duration="20" hangup="16"
auth_username="VP_ENV_USERNAME"
password="VP_ENV_PASSWORD"
realm="target.com"
rtp_stats="true"
>
<x-header name="X-Foo" value="Bar"/>
</action>
<!-- note: param value starting with VP_ENV_ will be replaced by environment variables -->
<!-- note: rtp_stats will include RTP transmission statistics -->
<!-- note: x-header tag inside an action will append an header. You can add any header like User-Agent with this method -->
<action type="wait" complete="true"/>
</actions>
</config>
Example: starting a TLS server
./voip_patrol \
--port 5060 \ # TLS port 5061 +1
--conf "xml/tls_server.xml" \
--tls-calist "tls/ca_list.pem" \
--tls-privkey "tls/key.pem" \
--tls-cert "tls/certificate.pem" \
--tls-verify-server \
<config>
<actions>
<!-- note: default is the "catch all" account,
else account as to match called number -->
<action type="accept"
match_account="default"
hangup="5"
play_dtmf="0123456789#*"
play="voice_ref_files/f.wav"
code="200" reason="YES"
ring_duration="5"
/>
<!-- DTMF will be sent using RFC2833 -->
<!-- note: wait for new incoming calls
forever and generate test results -->
<action type="wait" ms="-1"/>
</actions>
</config>
Example: accepting calls and checking for specific header with exact match or regular expression and no match on other
<config>
<actions>
<action type="accept"
match_account="default"
hangup="5"
code="200" reason="OK"
>
<check-header name="Min-SE"/>
<!-- Check that a header exists -->
<check-header name="X-Foo" value="Bar"/>
<!-- Check that a header exists and have a specific value -->
<check-header name="From" regex="^.*sip:\+1234@example\.com"/>
<!-- Check that a header exists and matches a specific regex -->
<check-header name="To" regex="^.*sip:\+5678@example\.com" fail_on_match="true"/>
<!-- Check that a header exists and NOT matches a specific regex -->
<check-header name="RURI" regex="^INVITE\ sip:\d{5}@(\d{1,3}\.){3}\d{1,3}:\d{1,5};.*transport=[a-zA-Z]{3};.*"/>
<!-- Not really a header, but allows to check the Request URI on an incoming INVITE-->
</action>
<action type="wait" ms="-1"/>
</actions>
</config>
Example: accepting calls and searching the message with a regular expression
<config>
<actions>
<action type="accept"
match_account="default"
hangup="5"
code="200" reason="OK"
>
<check-message method="INVITE" regex="m=audio(.*)RTP/AVP 0 8.*"/>
<!-- searching for pcmu pcma in the SDP -->
</action>
<action type="wait" ms="-1"/>
</actions>
</config>
Example: accepting calls and searching the message with a regular expression that should not be there
<config>
<actions>
<action type="accept"
match_account="default"
hangup="5"
code="200" reason="OK"
>
<check-message method="INVITE" regex="m=audio(.*)RTP/AVP 0 8.*" fail_on_match="true"/>
<!-- searching for pcmu pcma in the SDP, but this is wrong here -->
</action>
<action type="wait" ms="-1"/>
</actions>
</config>
Example: making tests calls with wait_until
Scenario execution is sequential and non-blocking. We can use “wait” command with previously set “wait_until” params to control parallel execution.
Call States
NULL : Before INVITE is sent or received
CALLING : After INVITE is sent
INCOMING : After INVITE is received.
EARLY : After response with To tag.
CONNECTING : After 2xx is sent/received.
CONFIRMED : After ACK is sent/received.
DISCONNECTED
<config>
<actions>
<action type="call" label="call#1"
transport="udp"
wait_until="CONFIRMED"
expected_cause_code="200"
caller="15148888888@noreply.com"
callee="12011111111@target.com"
/>
<!-- note: will wait until all tests pass wait_until state -->
<action type="wait"/>
<action type="call" label="call#2"
transport="udp"
wait_until="CONFIRMED"
expected_cause_code="200"
caller="15147777777@noreply.com"
callee="12012222222@target.com"
/>
<action type="wait" complete="true"/>
</actions>
</config>
Example: testing registration
<config>
<actions>
<!-- note: proxy param to send to a proxy -->
<action type="register" label="register target.com"
transport="udp"
account="VP_ENV_USERNAME"
username="VP_ENV_USERNAME"
auth_username="VP_ENV_USERNAME"
password="VP_ENV_PASSWORD"
proxy="172.16.7.1"
realm="target.com"
registrar="target.com"
expected_cause_code="200"
/>
<action type="wait" complete="true"/>
</actions>
</config>
Example: re-invite with new codec
<config>
<action>
<action type="codec" disable="all"/>
<action type="codec" enable="pcma" priority="250"/>
<action type="codec" enable="pcmu" priority="248"/>
<!-- call that will last 12 seconds and re-invite every 2 seconds -->
<action type="call"
wait_until="CONFIRMED"
expected_cause_code="200"
caller="16364990640@125.22.198.115"
callee="12349099229@sip.mydomain.com"
max_duration="55" hangup="12"
auth_username="65454659288" password="adaadzWidD7T"
realm="sip.mydomain.com"
re_invite_interval="2"
rtp_stats="true"
/>
<action type="wait"/> <!-- this will wait until the call is confirmed -->
<action type="codec" disable="pcma"/>
<!-- re-invite will now use pcmu forcing a new session -->
<action type="wait" ms="3000"/> <!-- this will wait 3 seconds -->
<action type="codec" enable="pcma" priority="250"/>
<!-- re-invite will now use pcma forcing a new session -->
<action type="wait" complete="true"> <!-- Wait until the calls are disconnected -->
<actions/>
<config/>
Example: Overwriting local contact header
<config><actions>
<action type="codec" disable="all"/>
<action type="codec" enable="pcma" priority="250"/>
<action type="codec" enable="gsm" priority="249"/>
<action type="codec" enable="pcmu" priority="248"/>
<action type="call"
transport="udp"
caller="+15147371787@fakecustomer.xyz"
callee="+911@edgeproxy1"
transport="udp"
auth_username="20255655"
password="qntzhpbl"
realm="sip.flowroute.com"
rtp_stats="true"
late_start="false"
force_contact="sip:+15147371787@10.10.2.5:5777"
play="/git/voip_patrol/voice_ref_files/reference_8000_12s.wav"
hangup="5">
<x-header name="Foo" value="Bar"/>
</action>
<action type="wait" complete/>
</actions></config>
Example: WAIT action
wait forever:
<action type="wait" ms="-1"/>
wait until you receive a certain amount of calls
<action type="accept" call_count="x" ... />
<action type="wait" complete="true"/>
wait 5 seconds or one call
<action type="accept" call_count="1" ... />
<action type="wait" ms="5000"/>
Sample JSON output RTP stats report with multiples sessions
one block is generated everytime a session is created
{
"rtp_stats_0": {
"rtt": 0,
"remote_rtp_socket": "10.250.7.88:4028",
"codec_name": "PCMA",
"clock_rate": "8000",
"Tx": {
"jitter_avg": 0,
"jitter_max": 0,
"pkt": 105,
"kbytes": 16,
"loss": 0,
"discard": 0,
"mos_lq": 4.5
},
"Rx": {
"jitter_avg": 0,
"jitter_max": 0,
"pkt": 104,
"kbytes": 16,
"loss": 0,
"discard": 0,
"mos_lq": 4.5
}
},
"rtp_stats_1": {
"rtt": 0,
"remote_rtp_socket": "10.250.7.89:40230",
"codec_name": "PCMU",
"clock_rate": "8000",
"Tx": {
"jitter_avg": 0,
"jitter_max": 0,
"pkt": 501,
"kbytes": 78,
"loss": 0,
"discard": 0,
"mos_lq": 4.5
},
"Rx": {
"jitter_avg": 0,
"jitter_max": 0,
"pkt": 501,
"kbytes": 78,
"loss": 0,
"discard": 0,
"mos_lq": 4.5
}
}
}
Example: email reporting
<config>
<actions>
<action type="alert"
email="jchavanton+vp@gmail.com"
email_from="test@voip-patrol.org"
smtp_host="smtp://gmail-smtp-in.l.google.com:25"
/>
<!-- add more test actions here ... -->
<action type="wait" complete="true"/>
</actions>
</config>
accept command parameters
Name | Type | Description |
---|---|---|
ring_duration | int | ringing duration in seconds |
expected_duration | int | expected duration of the call in seconds. Test considered failed if actual duration is different |
expected_setup_duration | int | expected duration of the call setup (INVITE - 200 OK) in seconds. Test considered failed if actual duration is different |
early_media | bool | if true 183 with SDP and early media is used |
timer | string | control SIP session timers, possible values are : inactive, optional, required or always |
code | int | SIP cause code to return must be > 100 and < 700 |
expected_cause_code | int | SIP cause to be expected from caller side as a call result. Value 487 could be combined with fail_on_accept parameter |
match_account | string | Account will be used to receive this call (made via register ) falling back to match the user part of an incoming call RURI or default will catch all.</br>Point, in this case account parameters specified at register will override account-specific parameters that defined here, for ex. transport or srtp |
response_delay | int | delay before 100 - Trying reponse is sent in seconds. Useful to test timeouts and race conditions |
call_count | int | The amount of calls to receive to consider the command completed, default -1 (considered completed) |
transport | string | Force a specific transport for all messages on accepted calls, default to all transport available |
force_contact | string | optional URI to be put as Contact for accept account. Helps bypass NAT-related issues during inbound call testing |
play | string | path to file to play upon answer |
record | string | path to file to record audio upon answer. Can be auto , in this case filename would be /srv/<call_id>_<remote_contact>_rec.wav |
record_early | bool | if true early media will be also recorded |
play_dtmf | string | list of DTMF symbols to be sent upon answer |
re_invite_interval | int | Interval in seconds at which a re-invite with SDP will be sent |
rtp_stats | bool | if true the json report will include a report on RTP transmission |
min_mos | float | Minimal MOS value for this call |
srtp | string | Comma-separated values of the following sdes - add SDES support, dtls - add DTLS-SRTP support, force - make SRTP mandatory |
cancel | string | optional - mark the test passed, if the call was canceled by the caller before answer, force - mark test passed ONLY if the call was canceled by the caller. Make sure that you set ring_duration > 0 |
fail_on_accept | bool | If true - than accepting this call counts as a failed test |
disable_turn | bool | If true - global turn configuration is ignored for this account |
hangup | int | call duration in second before hangup |
call command parameters
Name | Type | Description |
---|---|---|
timer | string | control SIP session timers, possible values are : inactive, optional, required or always |
proxy | string | ip/hostname of a proxy where to send the call |
caller | string | user@host , mandatory parameter (also used in the From header unless from is specified) |
from | string | From header complete "Display Name" <sip:test at 127.0.0.1> in a format "Display Name" <sip:test at 127.0.0.1> |
callee | string | request URI user@host (also used in the To header unless to_uri is specified) |
to_uri | string | used@host part of the URI in the To header |
auth_username | string | authentication username on INVITE |
password | string | password used on INVITE |
realm | string | realm use for authentication on INVITE. If empty - any auth realm is allowed |
transport | string | force a specific transport tcp , udp , tls , sips , wss |
contact_uri_params | string | string, that will be added to Contact URI as params |
play | string | path to file to play upon answer |
record | string | path to file to record audio upon answer. Can be auto , in this case filename would be /srv/<call_id>_<remote_contact>_rec.wav |
record_early | bool | if true early media will be also recorded |
play_dtmf | string | list of DTMF symbols to be sent upon answer |
re_invite_interval | int | Interval in seconds at which a re-invite with SDP will be sent |
rtp_stats | bool | if true the json report will include a report on RTP transmission |
min_mos | float | Minimal MOS value for this call |
srtp | string | Comma-separated values of the following sdes - add SDES support, dtls - add DTLS-SRTP support, force - make SRTP mandatory. Note, if you don't specify force , call would be made with plain RTP |
late_start | bool | if true no SDP will be included in the INVITE and will result in a late offer in 200 OK/ACK |
disable_turn | bool | If true - global turn configuration is ignored for this account |
force_contact | string | local contact header will be overwritten by the given string |
max_ring_duration | int | max ringing duration in seconds before cancel |
expected_duration | int | expected duration of the call in seconds. Test considered failed if actual duration is different |
expected_setup_duration | int | expected duration of the call setup (INVITE - 200 OK) in seconds. Test considered failed if actual duration is different |
hangup | int | call duration in second before hangup |
repeat | int | do this call multiple times |
register command parameters
Name | Type | Description |
---|---|---|
proxy | string | ip/hostname of a proxy where to send the register |
username | string | AOR username - From/To/Contact header user part |
auth_username | string | authentication username, account name, From/To/Contact header user part. If not specified, username is used |
password | string | account password |
account | string | if not specified username is used. Internal identifier, also used in match_account in accept action |
aor | string | Account Address Of Record. if not specified - <usename@registrar> |
contact_uri_params | string | string, that will be added to Contact URI as params |
registrar | string | SIP UAS handling registration where the messages will be sent |
transport | string | force a specific transport tcp , udp , tls , sips , wss |
realm | string | realm use for authentication. If empty - any auth realm is allowed |
srtp | string | Comma-separated values of the following sdes - add SDES support, dtls - add "DTLS-SRTP" support, force - make SRTP mandatory. Used for incoming calls to this account |
disable_turn | bool | If true - global turn configuration is ignored for this account. Used for incoming calls to this account |
unregister | bool | unregister the account <usename@registrar;transport=x> |
reg_id | int | if present outbound and other related parameters will be added (see RFC5626) |
instance_id | int | same as reg_id , if not present, it will be generated automatically |
rewrite_contact | bool | default true , detect public IP when registering and rewrite the contact header |
message command parameters
Name | Type | Description |
---|---|---|
from | string | From header complete ""Display Name" <sip:test at 127.0.0.1>" |
to_uri | string | used@host part of the URI in the To header |
transport | string | force a specific transport <tcp,udp,tls> |
realm | string | realm use for authentication. If empty - any auth realm is allowed |
username | string | authentication username, account name, From/To/Contact header user part |
password | string | authentication password |
label | string | test description or label |
Example: sending a message
<?xml version="1.0"?>
<config>
<actions>
<action type="message" label="testing SIP message" transport="udp"
expected_cause_code="202"
text="Message in a bottle."
from="123456@in.the.ocean"
to_uri="15876580542@in.the.ocean"
username="123456"
password="pass"
/>
<action type="wait" complete="true"/>
</actions>
</config>
accept_message command parameters
Name | Type | Description |
---|---|---|
account | string | Account will be used if it matches the user part of an incoming message RURI or "default" will catch all |
message_count | int | The amount of messages to receive to consider the command completed, default -1 (considered completed) |
transport | string | Force a specific transport for all messages on accepted messages, default to all transport available |
label | string | test description or label |
Example: receiving a message
<?xml version="1.0"?>
<config>
<actions>
<action type="register" label="register" transport="udp"
expected_cause_code="200"
username="123456"
password="password"
registrar="pbx.somewhere.time"
/>
<action type="wait" complete="true"/>
<action type="accept_message"
account="123456"
message_count="1"
/>
<action type="wait" complete="true"/>
</actions>
</config>
wait command parameters
Name | Type | Description |
---|---|---|
complete | bool | if true wait for all the test to complete (or reach their wait_until state) before executing next action or disconnecting calls and exiting, needed in most cases |
ms | int | the amount of milliseconds to wait before executing next action or disconnecting calls and exiting, if -1 wait forever |
Example: codec configuration
<config>
<actions>
<action type="codec" disable="all"/>
<action type="codec" enable="pcmu" priority="250"/>
<!-- more actions ... -->
<action type="wait" complete/>
</actions>
</config>
codec command parameters
Name | Type | Description |
---|---|---|
priority | int | 0-255, where zero means to disable the codec |
enable | string | Codec payload type ID, ex. "g722", "pcma", "opus" or "all" |
disable | string | Codec payload type ID, ex. "g722", "pcma", "opus" or "all" |
Example: TURN configuration
<config>
<actions>
<action type="turn" enabled="true" server="x.x.x.x:3478" username="foo" password="bar"/>
<!-- more actions ... -->
<action type="wait" complete/>
</actions>
</config>
turn command parameters
Name | Type | Description |
---|---|---|
enabled | bool | if "true" STUN/TURN/ICE server usage will be enabled |
server | string | STUN/TURN server URI or IP:port |
username | string | TURN server username |
password | string | TURN server password |
password_hashed | bool | if "true" use hashed password, default plain password |
sip_stun_use | bool | if "true" SIP reflective IP is use with signaling |
media_stun_use | bool | if "true" STUN reflective IP is use with media/SDP |
stun_only | bool | if "true" TURN and ICE are disabled and only STUN is use |
disable_ice | bool | if "true" ICE mechanism is disabled |
ice_trickle | bool | if "true" Trickle ICE mechanism is used |
using env variable in scenario actions parameters
Any value starting with VP_ENV
will be replaced by the envrironment variable of the same name.
Example : username="VP_ENV_USERNAME"
export VP_ENV_PASSWORD=????????
export VP_ENV_USERNAME=username
Docker
voip_patrol/docker$ tree
.
├── build.sh # docker build command example
├── Dockerfile # docker build file for Linux Alpine
└── voip_patrol.sh # docker run example starting
Dependencies
PJSUA2
PJSUA2 : A C++ High Level Softphone API : built on top of PJSIP and PJMEDIA http://www.pjsip.org http://www.pjsip.org/docs/book-latest/PJSUA2Doc.pdf
External tool to test audio quality
PESQ
P.862 : Perceptual evaluation of speech quality (PESQ): An objective method for end-to-end speech quality assessment of narrow-band telephone networks and speech codecs http://www.itu.int/rec/T-REC-P.862
./run_pesq +16000 voice_files/reference.wav voice_files/recording.wav