detailed_fetch_error.enable
Section: vxmli
Default Value: false
Valid Values: Choose between: true or false
Changes Take Effect: At start/restart
Specifies whether a more detailed fetch error should be generated. If enabled, a different error id
and trap is generated for fetching errors of type 4xx, all other fetching errors will be handled as usual. In addition,
the CallId is included to all fetching error log message.
enable_dns_cache
Section: sip
Default Value: false
Valid Values: Choose between: true or false
Changes Take Effect: At start/restart
Specifies if MCP should enable or disable the use of DNS cache. Enabling DNS cache increases MCP's resilience towards network issues between MCP and the DNS Servers. If the option is enabled, the target address is retrieved from the DNS cache, if available. If unavailable, a fresh DNS query will be used to retrieve the target address and the result will be cached depending on the DNS query response. If the option is disabled, target address is resolved by a fresh DNS query.
opus.fmtp
Section: mpc
Default Value: useinbandfec=1
Valid Values: [useinbandfec=(0|1)] [; maxaveragebitrate=(6000-510000)]
Changes Take Effect: At start/restart
Specifies the default OPUS fmtp option string used by the platform when initiating SDP offer or answering SDP offer. If the fmtp parameter is empty, the fmtp attribute line for OPUS will not be present in SDP offered by the platform.
The supported fmtp parameters are:
"useinbandfec" specifies whether MCP will take advantage of Opus in-band FEC or not. Opus in-band FEC helps mitigate errors when unstable transmissions occur. Possible values for "useinbandfec" are 1 and 0. If no value is specified, then in-band FEC will not be applied. For better error recovering results, the corresponding "maxaveragebitrate" should be set to a higher value, e.g. 48000.
"maxaveragebitrate" specifies the maximum average bitrate. Sending out this ftmp serves as a mandate to the remote end to not send a stream with average birate higher than the "maxaveragebitrate" value, as it might overload our network and/or receiver. Possible values for "maxaveragebitrate" are any integer in the range from 6000 to 510000 inclusive. If no value is specified, the bitrate will be automatically decided by MCP according to the audio source.
Unsupported fmtp parameters: usedtx; cbr; stereo; sprop-stereo; maxplaybackrate; sprop-maxcapturerate.
opus.packetloss
Section: mpc
Default Value: 20
Valid Values: A number between 0 and 100 inclusive.
Changes Take Effect: Immediately/session
Used to indicate the expected packet loss percentage to the Opus encoder. The range is 0 to 100, inclusive. Higher values will give progressively more loss resistant behavior, at the expense of quality at a given bitrate in the lossless case, but greater quality under loss. The default value is 20, implying 20 percent packet loss is expected.
opus.complexity
Section: mpc
Default Value: 10
Valid Values: A number between 0 and 10 inclusive.
Changes Take Effect: Immediately/session
Specifies Opus encoder's computational complexity. The range is 0 to 10, inclusive. The higher the value, the higher the complexity and the audio quality.
opus.apptype
Section: mpc
Default Value: voice
Valid Values: voice: VoIP Application that mostly uses speech. It may alter the output sound to make it more intelligible.
audio: Application that uses music and speech. It provides lower coding delay.
Changes Take Effect: Immediately/session
Used to indicate the intended application type to the Opus encoder.
opuspayload
Section: mpc
Default Value: 116
Valid Values: A valid Opus payload can only be an integer from 96 to 127 inclusive
Changes Take Effect: At start/restart
Default payload type number to use for the Opus codec
cpa.postconnsilduration
Section: mpc
Default Value: 1000
Valid Values: A valid value must be in the range 200-3000 inclusive
Changes Take Effect: Immediately/session
Specifies the duration of silence that may trigger a configurable CPD result immediately after the postconnent event arrives. The default value is 1000 milliseconds. The range is between 200 and 3000 milliseconds. This parameter does not take effect until mpc.cpa.postconnsilresult is set to a value other than none (0).
cpa.postconnsilresult
Section: mpc
Default Value: 0
Valid Values: Choose between: 0, 2, 15, 16, 17 or 18.
Changes Take Effect: Immediately/session
Specifies CPA result if a period of silence, which is configurable through mpc.cpa.postconnsilduration, is detected immediately after postconnect event arrives.
The default option is none (0), that is, no CPA result returned.
0 - none
2 - busy
15 - human
16 - machine
17 - fax
18 - no_media
curlconnecttimeout
Section: fm
Default Value: 300
Valid Values: An positive integer less than equal to 65535.
Changes Take Effect: At start/restart
Specifies the maximum time in seconds that is allowed for the connection phase to the server. This value applies only to the
connection phase; it has no effect once the connection is made.
Note: For Netann calls, the parameter annc.fetchtimeout has a maximum value of 25 seconds. This can limit the maximum value
for curlconnecttimeout. For example, if annc.fetchtimeout is set to 25 seconds and curlconnecttimeout is set to 30 seconds,
the call will terminate as soon as the annc.fetchtimeout timer expires. Similarly, for VXML calls, parameters
sessmgr.acceptcalltimeout and vxmli.initial_request_fetchtimeout might limit the maximum value for curlconnecttimeout.
8.5.185.34
Voice Platform Media Control Platform Release Notes
Release Date | Release Type | Restrictions | AIX | Linux | Solaris | Windows |
---|---|---|---|---|---|---|
01/22/18 | General | X | X |
Helpful Links
Releases Info
Product Documentation
Genesys Products
What's New
This release contains the following new features and enhancements:
- Media Control Platform (MCP) now supports configuration of the HTTP and HTTPS connection timeout through the parameter curlconnecttimeout. This parameter specifies the maximum time in seconds before the HTTP/HTTPS connection attempt times out. This value applies only to the connection phase, and has no effect once the connection is made.
- MCP is no longer blocked while waiting for an SSL connection to a web server through an HTTPS proxy to finish.
- MCP introduces two new configuration options, cpa.postconnsilresult and cpa.postconnsilduration, to return a configurable CPA result if a configurable length of silence is detected right after the postconnect event arrives.
- As part of the Cloud requirement, custom prompts under the /$installation location/music/on_hold_saas folder are now available as a part of MCP installation package.
- MCP now supports OPUS audio codec for RTP transport along with transcoding to and from OPUS. Recording and playback using OPUS are currently not supported. The following are the new MCP configuration parameters:
- MCP now displays the new Genesys logo for video conference participants who do not provide video input.
- MCP now supports internal caching of DNS A/AAAA queries for SIP interfaces. This feature can be enabled through the parameter enable_dns_cache. By default, DNS queries are not cached.
Resolved Issues
This release contains the following resolved issues:
MCP now sends the RTSP DESCRIBE message when it connects to the MRCP Proxy regardless of when the MRCP proxy starts. (GVP-35379)
MCP play cache is now cleaned up during MCP installation. (GVP-35375)
MCP no longer crashes when a remote party fails to respond to a REFER request and MCP fails to respond to the BYE message with 200 OK. (GVP-35237)
MCP no longer crashes when a transfer is started immediately after a previous, unsupported transfer fails and an error.unsupported.transfer event occurs. (GVP-35143)
MCP now sends a detailed SNMP trap related to fetching errors. If the feature is enabled, MCP sends a new trap for HTTP 4xx failures. The old trap is still used for all other HTTP failures regardless of whether the feature is enabled or disabled. Moreover, the Call-ID is now added to both the new and existing traps. The feature can be enabled by setting the parameter detailed_fetch_error.enable to true. (GVP-34553)
MCP no longer terminates when the RTCP port is explicitly offered in the SDP message. (GVP-24168)
MCP now accepts the transport offered when MRCP server changes the transport in the Contact header of the 200 OK response. Previously, MCP did not accept the change and responded with the same transport as in the INVITE message. (GVP-24140)
The MCP process no longer deadlocks when it is terminating a recording session and a response to a Re-INVITE message comes in at the same time. (GVP-22802)
Upgrade Notes
No special procedure is required to upgrade to release 8.5.185.34.