Accepted Solutions
I think this document explains why you can't setup a VC bridge. It been updated in October 2012, so it's really resent. http://www.cisco.com/en/US/prod/collateral/iosswrel/ps6537/ps6555/ps6602/whitepaper_C11-718333.pdf
According to it you can't have a fragmented DSP core, aka use the same core for both voice and video. Each of your PVDM3 modules have one core and if you use 40 channels of the total 48 for voice you would use both cores for voice, therefor you can't allocate the remaining 8 channels to the VC bridge.
If I calculate 34% of 48 I get just above 16. So at least 16+1 channel is allocated by the voice-service dsp-reservation 34 command.
Please remember to rate all useful posts.
Sent from Cisco Technical Support iPhone App
How do you figure you have enough DSPs? 4 BRI cards by themselves will use the PVDM3-32, here is what I get with the calc:
This is for 1 video bridge -4 participants - 1 class - homogeneous
Hi Chris,
I tried so set a lower value for "voice-service dsp-reservation", but 34 was the lowest possible value.
As far as i know, the information that, DSP cannot be shared betwen TDM and conference ressources is not valid anymore for PVDM3, since a new firmware is used, that provides all services:
Q.
Can conferencing, transcoding and voice calls be supported on a single DSP?
A. Yes. The PVDM3 modules have a universal firmware image that allows sharing DSP resources between transcoding, voice, and conference calls. On the PVDM2 you can use the same DSP for voice and transcoding calls, but a different DSP firmware image is required for conference calls requiring a DSP to be dedicated when conferencing is configured.
Anyway: my BRI-cards do not use the full PVDM3-32 and the PVDM3-16 should still be enough for one video-conference-profile:
ABout the audio-codecs: IOS installed them by itself when I created the dspfarm-profile.
Regards
Florian
Even if a lot of codecs get automatically added once you do a conference dspfarm-profile you can manually remove them. The profile need to be shutdown and can't have any max session statement to be able to remove them.
All g729 codecs with a "b" in the name should always be removed if you don't need to use VAD. A golden rule is VAD is BAD.
Please remember to rate all useful posts.
Sent from Cisco Technical Support iPhone App
I've never setup a VC conf bridge on a G2 ISR, so I have no first had info on how to do it. But I found this post with a little digging with Google. http://www.ccierants.com/2011/12/cisco-isr-g2-video-conferencing.html?m=1
Take a look at it and report back if there were any useful info in it.
Please remember to rate all useful posts.
Sent from Cisco Technical Support iPhone App
In the link on the blog I saw this explanation of when you see a 0 (zero) after the max session command.
Q. Why do I get maximum-sessions 0 after configuring the required codecs?
A. Display of maximum-sessions 0 while configuring your dspfarm indicates that not enough DSP resources are available. You can try to remove certain codecs from your profile or add additional DSP resources. Please use the DSP calculator to calculate the number of DSPs you need.
To me that indicate that you either don't have enough DSPs or need to remove some codecs from the dsp profile. All g729 with b in the name are a higher complexity codecs than the other. So these would require more dsp resources. Also the minimum amount you can setup for video based on the FAQ seem to be 4, check with the dsp calculator that you have enough PVDM3 modules for this.
Please remember to rate all useful posts.
Sent from Cisco Technical Support iPhone App
I think this document explains why you can't setup a VC bridge. It been updated in October 2012, so it's really resent. http://www.cisco.com/en/US/prod/collateral/iosswrel/ps6537/ps6555/ps6602/whitepaper_C11-718333.pdf
According to it you can't have a fragmented DSP core, aka use the same core for both voice and video. Each of your PVDM3 modules have one core and if you use 40 channels of the total 48 for voice you would use both cores for voice, therefor you can't allocate the remaining 8 channels to the VC bridge.
If I calculate 34% of 48 I get just above 16. So at least 16+1 channel is allocated by the voice-service dsp-reservation 34 command.
Please remember to rate all useful posts.
Sent from Cisco Technical Support iPhone App
A. Yes. The PVDM3 modules have a universal firmware image that allows sharing DSP resources between transcoding, voice, and conference calls.
This is only my interpretation of this and the other links in this post, so I can be wrong, but anyway here goes what I think is the problem in your case.
The part you reference from in the FAQ “talks” about the ability for PVDM3 modules to handle mixed voice resources, as transcoding, TDM to/from VoIP, and conference would be. But in the link I referenced it “talks” about the ability to mix voice and video resources on PVDM3 modules. See below snipplet from the document I linked to previously.
IPSLA VO presently cannot use fragmented cores. A fragmented core is one that is allocated for both voice and video. This needs to be taken into consideration when procuring/provisioning PVDM3.
But a TAC case would be of interest, so that we can get an official take from Cisco on this. Please update when you have got an answer from TAC.
Please remember to rate helpful responses and identify helpful or correct answers.
