Media server specific parameters

The following server options are controlled via the media_server.conf file. To change the media server configuration settings, you can edit media_server.conf that is found in mrcp-api/docker/lumenvox/ when MRCP is installed in the Linux based docker environment.

See Configuring Parameters for more information about changing various Media Server parameters.

Please make sure to read the information on the mrcp_server_ip setting below.

mrcp_server_ip

This is the only parameter that you really NEED to set. All others have acceptable defaults. This number should always be set to the IP address of the machine running the Media Server itself. This is the IP address that the Media Server puts in the SIP/MRCP packets that it sends to the client.

Only leave this default loopback IP address if you install the MRCP server on the same machine as the platform that is hosting your application.

Default: 127.0.0.1

mrcp_server_port_base

The lowest numbered port that will be used for MRCP. One port will be used per MRCP resource.

Note that this value may need to be changed to avoid any conflict with other applications and also operating system ephemeral ports

Default: 30000

rtp_server_port_base

The lowest numbered port that will be used for RTP and RTCP. Two sequentially numbered ports will be used per resource, one for RTP and the next for RTCP.

Note that this value may need to be changed to avoid any conflict with other applications and also operating system ephemeral ports

Default: 35000

mrcp_server_resmax

The maximum number of concurrent resources. Practically speaking, this number cannot be greater than the number of port licenses you have for the Speech Engine. However, you can have this number of RTP ports and also this number of MRCP ports, providing that license requirements are met.

Note that you may need to increase this value to account for all ASR and TTS resources you will be using in the Media Server, otherwise all of the resources may not be available. 

Default: 200

monitoring_port

Note: this setting is deprecated as of LumenVox legacy version 10.4

Sets the port on which the Media Server will listen for incoming client connections from the Media Server's monitoring tool.

Default: 29900

messaging_port

Sets the port on which the Media Server will listen for incoming LVManager messages

Default: 7590

sre_ip

If you are running the Media Server and the Speech Engine on different machines, this parameter should be the IP address of the machine running the Speech Engine. You may specify multiple speech servers by providing multiple IP addresses separated by semicolons.

Default: 127.0.0.1

enable_sre_logging

The LumenVox Speech Engine can save response files (files with the extension .callsre) that contain data about each interaction, including recognition results and audio. This parameter determines whether those files are created and what level of information is stored in them.

Note that, when enabled, these .callsre files can take up a lot of disk space if not maintained

Possible Values:

ValueMeaning
0SAVE_SOUND_FILES_NONE
Deactivates saving of .callsre files.
1SAVE_SOUND_FILES_BASIC
Saves basic information in .callsre files. This includes the audio from when BARGE_IN occurred up to the END_OF_SPEECH.
2SAVE_SOUND_FILES_ADVANCED
In addition to the information stored with the SAVE_SOUND_FILES_BASIC setting, when speech is streamed in to the Voice Activity Detection module, all of the collected data prior to a StreamCancel command is saved. This is usually when a NO_INPUT or TIMEOUT event occurs. This option collects data only when there is something to debug, which offers a balance between disk usage and retaining important information for debugging and is particularly useful when diagnosing NO-INPUT problems.
3SAVE_SOUND_FILES_ALL
Used to collect all streamed data in all cases. This option collects all streamed data, whether there was a NO_INPUT or TIMEOUT event or not, including untrimmed audio from 'good' decodes. This option can be useful in diagnosing barge-in problems as well as other potential issues.


Default : 0

enable_logging

Note: this setting is deprecated as of LumenVox  legacy version 10.4, replaced with a more flexible logging_verbosity setting in client_property.conf

Enables logging to the media_server_log file. 0 disables and 1 enables

Default : 1

enable_app_stat_logging

Updates statistics file when enabled. 0 disables and 1 enables

The statistics file will track call volume and performance over time. This file is named media_server_status.txt in the logs folder.

Default : 1

listening_socket_size

Sets the number of sockets that are allowed to backlog on the listen function. Reduce this number if your OS cannot handle the default value.

A higher number here allows better call surge performance.

Default : 150

port_allocation_mode

Sets the mode used for MRCP/RTP port allocation.

0 - ROUND-ROBIN approach, which may be optimal in large configurations
1 - FIRST-AVAILABLE port uses lowest port number available..

Default : 0

num_spawning_threads

Number of threads dedicated to handling RTSP and SIP call creation.

Note that if you are planning to operate a large number of calls through the Media Server, and would like to allow for handling large spikes of inbound traffic, increasing this value is recommended.

For example, if you anticipate times where there may be more than 200 calls simultaneously started (not total calls), you should set this value to 100. This value should be approximately 50% of the peak number of simultaneous new calls being handled. Increasing this value will allocate more system resources, so use your best judgment for the amount of traffic you anticipate.

Default : 10

num_graveyard_threads

Number of threads performing call cleanup.

This value seldom needs to be increased, however if you have an unusually and perpetually high call volume, increasing this number will help clean up resources after call sessions have ended.

Default : 10

num_mrcp_event_threads

Number of MRCP processing threads.

There should not be a need to modify this setting unless directed to do so by LumenVox support staff

Default : 24

num_rtp_event_threads

Number of threads handling RTP traffic.

There should not be a need to modify this setting unless directed to do so by LumenVox support staff

Default : 12

sip_port

The port on which the server will listen for incoming SIP requests.

Setting this to a value of 0 will disable the SIP listening port (along with SIP functionality) and potentially free up some resources if not needed.

Note that is you are running multiple SIP servers on the same machine, you may need to alter this setting so that they do not conflict with each other

Default: 5060

rtsp_port

The port on which the server will listen for incoming RTSP requests. 

Setting this to a value of 0 will disable the RTSP listening port (along with RTSP functionality) and potentially free up some resources if not needed.

Default: 554

max_num_rtp_packets_buffered

The maximum number of unprocessed RTP packets held on to when the Media Server is not in recognition mode. These packets are spooled in whenever a new recognition starts.

Do not use if the Media Server session is being shared with different calls since the noise baseline will get created with this spooled data for potentially another call.

Default: 0

Encryption Settings

sips_port

The port the Media Server will listen for new MRCP v2 sessions (using the SIPS protocol).. 

Setting this to a value of 0 will disable the SIP listening port (along with SIPS functionality) and potentially free up some resources if not needed.

Note that is you are running multiple SIP servers on the same machine, you may need to alter this setting so that they do not conflict with each other.

Default: 5061

sips_ssl_cert_file

SSL certificate (*.pem) file to use with SIPS. 

This should be a full path and filename, the file needs to be on the same server as the media server.

Default: self-signed certificate created at installation

sips_cipher_list

List of allowed ciphers to use during SIPS SSL/TLS negotiation.. 

Default: All valid ciphers

mrcp_tls_cert_file

SSL public key certificate (*.pem) file to use with MRCP v2 TLS. 

This should be a full path and filename, the file needs to be on the same server as the media server.

Default: self-signed certificate created at installation

mrcp_tls_cipher_list

List of allowed ciphers to use during MRCP v2 SSL/TLS negotiation.. 

Default: All valid ciphers

allow_unauthenticated_srtp

Controls if SIPS sessions that specify UNAUTHENTICATED_SRTP are allowed  . 

Default: 1 (Allow)

allow_unencypted_srtp

Controls how SIPS sessions that specify UNENCRYPTED_SRTP are handled. 

New MRCP v2 sessions that specify UNENCRYPTED_SRTP, will encrypt the SIP and MRCP messages, but not the audio.

By default, the media server will accept these sessions, to disable allowing un-encrypted audio with SIPS, set this to 0.

Default: 1 (Allow)


Was this article helpful?
Copyright (C) 2001-2024, Ai Software, LLC d/b/a LumenVox