10-02-2014 08:19 AM - edited 03-17-2019 12:24 AM
Hi there,
I am having an issue streaming multicast music on hold at remote sites by sourcing the music on hold from the local flash of the Cisco 2951.
I am running CUCM 9.1(2) SU1, I have 1 x PUB and 1 x SUB. I have enabled the PUB to stream unicast MOH and have enabled multicast MOH on the SUB. Base IP: 239.1.1.1 and Port 16384.
I have set the MOH Region to all remote sites regions to G.711 and have put the SUB into an MRG and have enabled MOH at the MRG and Audio Source Level as per documentation. The MRGL for remote sites contain the local sites hardware resources such as MTP, CFB and XCODE. Second to that MRG, I have set the MOH_MRG containing the SUB enabled for multicast.
When a remote site user put the PTSN caller on hold, I can see an active multicast music on hold session active on the local router but the PSTN caller hears silence !!! The number of In/Out packet on the multicast session does not increase at all.
Could you please help me guys? I have tried all options and still I am getting silence all the times.
Below is the configuration on the local gateway at remote sites:
XX-XXX-VG01#sh run | s call-m
call-manager-fallback
max-conferences 4 gain -6
transfer-system full-consult
ip source-address 10.114.80.1 port 2000
max-ephones 58
max-dn 120
transfer-pattern 031580....
keepalive 10
call-forward pattern 031580....
moh "music-on-hold.au"
multicast moh 239.1.1.1 port 16384 route 10.114.80.1 1.1.1.1
time-zone 29
time-format 24
date-format dd-mm-yy
Please note that when the PSTN caller is on hold, when I issue the # show ccm-manager music-on-hold, I get the following output all the times:
XX-XXX-VG01#sh voice call status
CallID CID ccVdb Port Slot/DSP:Ch Called # Codec MLPP Dial-peers
0x17740 22AA 0x3D1D09A4 0/0/0:15.4 0/1:2 4312 g711ulaw 200/202
1 active calls found
XX-XX-VG01#sh ccm-manager music-on-hold
Current active multicast sessions : 1
Multicast RTP port Packets Call Codec Incoming
Address number in/out id Interface
===================================================================
239.1.1.1 16384 0/0 96064 g711ulaw
XX-XXX-VG01#sh ephone summary
hairpin_block:
Max 58, Registered 0, Unregistered 0, Deceased 0 High Water Mark 59, Sockets 0
ephone_send_packet process switched 0
Max Conferences 32 with 0 active (4 allowed)
-------------------------------------
Skinny Music On Hold Status - group 0
Active MOH clients 0 (max 210), Media Clients 0, B-ACD Clients 0
File music-on-hold.au (not cached) type AU Media_Payload_G711Ulaw64k 160 bytes
Moh multicast 239.1.1.1 port 16384 route 10.114.80.1 1.1.1.1
I can confirm that the flash of the local router has indeed the music-on-hold.au file
XX-XXX-VG01#dir flash:
Directory of flash0:/
247 -rw- 496521 Sep 20 2014 10:30:06 +02:00 music-on-hold.au
263 -rw- 90063932 Sep 22 2014 11:24:06 +02:00 c2951-universalk9-mz.SPA.152-4.M6a.bin
256503808 bytes total (141819904 bytes free)
Solved! Go to Solution.
10-07-2014 10:31 AM
The Cisco developers asked us to upgrade to IOS 15.4(3)M. We have done that 10 and half hours ago. The multicast music on hold is working and the voice gateway has not crashed. We will continue to monitor.
10-07-2014 12:04 PM
Hi Josh,
Thanks so much for the info.
I have a change logged for Wednesday and Thursday. I will load the IOS Image and test as well. Will send you my feedback later on Wednesday around 19:00 GMT.
JK.
12-18-2014 07:37 PM
Hi Joshua
I am trying to set up multicast for music on hold on the cucm 9.
on the cucm, I enable multicast audio sources on MOH server, using base IP 239.1.1.1, and set increment multicast on " ip address"
however the multicast IP address keep changing, sometimes its 239.1.1.1, then after a while it suddenly change to 239.1.1.2 etc..
here is the config on our srst router ( which specify the multicast ip is 239.1.1.1):
is there any way that the multicast ip do not keep changing as its affecting the MOH at the moment.
" config for MOH"
telephony-service
video
srst mode auto-provision none
srst dn template 1
srst dn line-mode dual
max-ephones 265
max-dn 600
ip source-address 10.190.0.1 port 2000
timeouts interdigit 5
system message "GlobalDial Unavailable"
cnf-file location flash:
time-zone 48
date-format dd-mm-yy
voicemail 080758831#
max-conferences 8 gain -6
moh "xmas_jazz.wav"
multicast moh 239.1.1.1 port 16384 route 192.168.1.1
transfer-system full-consult
transfer-pattern 0.T
secondary-dialtone 0
create cnf-files version-stamp 7960 Nov 27 2013 18:23:59
I was thinking to put a few more multicast IP,. but it only allow you to put one ( while the cucm set the IP address as increment)
OC-MHO-VG-SRST(config)#telephony-service
OC-MHO-VG-SRST(config-telephony)#multi
OC-MHO-VG-SRST(config-telephony)#multicast ?
moh Set IP multicast address for use with music-on-hold from flash
OC-MHO-VG-SRST(config-telephony)#multicast moh ?
A.B.C.D Define music-on-hold IP multicast address from flash
OC-MHO-VG-SRST(config-telephony)#multicast moh 239.1.1.2 ?
port Define media port for multicast moh
OC-MHO-VG-SRST(config-telephony)#multicast moh 239.1.1.2
12-22-2014 01:54 PM
Hi Rachel,
The IP address CUCM selects is based on the codec and the audio stream.
See the following URL. http://voiceonbits.com/2010/06/29/moh-issues-and-resolution/
I borrowed this table from that URL and cleaned up the port numbers. Hopefully this helps.
Inc. Multicast on IP Address | Inc. Multicast on Port Number | ||||
Audio Stream | Codec | Dst. IP Address | Dst. Port | Dst. IP Address | Dst. Port |
1 | G.711 ulaw | 239.1.1.1 | 16384 | 239.1.1.1 | 16384 |
1 | G.711 Alaw | 239.1.1.2 | 16384 | 239.1.1.1 | 16386 |
1 | G.729 | 239.1.1.3 | 16384 | 239.1.1.1 | 16388 |
1 | Wideband | 239.1.1.4 | 16384 | 239.1.1.1 | 16390 |
2 | G.711 ulaw | 239.1.1.5 | 16384 | 239.1.1.1 | 16392 |
2 | G.711 Alaw | 239.1.1.6 | 16384 | 239.1.1.1 | 16394 |
2 | G.729 | 239.1.1.7 | 16384 | 239.1.1.1 | 16396 |
2 | Wideband | 239.1.1.8 | 16384 | 239.1.1.1 | 16398 |
10-02-2014 01:13 PM
Is the PSTN call using that same router as the voice gateway. If so, make sure you have "ccm-manager music-on-hold" enabled globally so it can listen to the multicast stream.
Is this an H.323 gateway or MGCP?
10-02-2014 01:18 PM
Hi Brian,
The Voice Gateway is H.323 and the E1 Controller is terminated on the same Voice Gateway. I am having the same problem at two different SRST sites. Once is running Cisco 2921 and the other is using a 2951, all are using E1 PRI.
The funny thing is that the only time I get a chance to play music on hold sourcing from the flash card of the Voice Gateway is when the site is operating in SRST Mode.
NOTE: I have applied the global config command # ccm-manager music-on-hold
Warm regards,
JK.
10-02-2014 01:22 PM
Since it's H.323, can you grab the follow debugs for a call placed on hold:
debug h225 asn1
debug h245 asn1
You'll want to disable console logging and monitor logging and use a buffer to get the debug output.
no logging console
no logging monitor
logging buffer 5000000 debug
service sequence-numbers
Do a clear log, enable the debugs, then run "show log" to see the output. I want to see if CallManager is telling the gateway to listen to the right multicast IP/port.
10-02-2014 01:28 PM
Thanks Brian,
I have already confirmed this with the Cisco TAC Engineer and as posted in my initial post, when the PSTN Caller is on hold, you can clearly see that the CUCM is telling the gateway to listen to the right multicast IP Address and Port number.
Below is the output of # show ccm-manager music when a PSTN Caller is put on hold.
PM-DBN-VG01#sh ccm-manager music-on-hold
Current active multicast sessions : 1
Multicast RTP port Packets Call Codec Incoming
Address number in/out id Interface
===================================================================
239.1.1.1 16384 0/0 96064 g711ulaw
PM-DBN-VG01#
As you can see above that all looked correct, except that there is silence. The Packets count in/out is not increasing at all and I do not see the incoming interface being populated at all.
When I look at another cluster which works well, I can see that the Packets In/Out do increase and the incoming interface is a Loopback interface when I put a PSTN caller on hold.
Any advise?
10-02-2014 01:35 PM
Yea, that part should be good then.
I'd try playing around with this config:
multicast moh 239.1.1.1 port 16384 route 10.114.80.1 1.1.1.1
What interface is 10.114.80.1 and 1.1.1.1? Might be worth setting up a loopback and adding that in the route statement.
10-02-2014 01:38 PM
IP 10.114.80.1 is the IP Address configured on the Gig 0/0 of the Cisco 2921 and 1.1.1.1 is a Loopback IP Address.
I have tried changing the loopback IP address, play with the route statement and so on but still getting the pure silence and nothing streaming at all.
I am not sure what I am missing this time around. I have done this kind of setup many times but this ones is puzzling for a while now.
:(
10-02-2014 01:43 PM
Do you have multicast-routing enabled? I've seen that cause issues before.
What does "show ip mroute" look like during a hold session?
10-02-2014 01:46 PM
IP Multicast Routing is not enabled on the Voice Gateway at all.
I have tried enabling that during off hours and had the same problem. Still hearing silence.
I have couple of CUCM Clusters with the same exact setup with hundred of SRST Sites and IP Multicast Routing is not enabled on those Gateways at all but everything work so perfectly.
10-02-2014 01:50 PM
You could also try manually subscribing your Gig0/0 interface to stream to see if the router is even streaming it at all:
ip igmp join-group 239.1.1.1
10-02-2014 01:53 PM
When the PSTN Caller is on hold and I issue the command # show ip igmp group, I can see that the Router is Joining the Group at 239.1.1.1
10-02-2014 01:47 PM
Also I think your moh statement is wrong:
moh "music-on-hold.au"
I don't think I usually put the quotes on it. Try:
no moh "music-on-hold.au"
moh music-on-hold.au
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: