I'd like to get a clarification of how DSP resources are allocated in the 2811. If we have a PVDM2-32, there are two DSPs available, as I understand it.
I read a note on CCO that indicated that a DSP can either be assigned to call termination or it can be used for conferencing, but not both. Is that correct?
If so, that means that I could have up to 8 G.729 calls terminating on the router as well as up to 8 conference participants. Is that correct?
Also, let's assume that I have no conference participants at the moment, but there are 8 calls terminating on the router. What happens to that 9th call? It sounds like it can't make use of the idle channels on the second DSP.
Let me know if I have this straight. It's a bit confusing. :)
That is correct, a dsp being used for voice call termination cannot be used as a conference resource. Alot depends on your topology and how many pstn lines you need to support. Typically, you would not be using g.729 from your local phones to your gateway, you would use g.711 which would decrease the codec complexity and increase the number of channels your dsp's could handle. There is a handy DSP calculator located here:
andy - berbee
Typically, most of our calls would be interoffice calls that go across the WAN. Is it a simple thing to make the calls downshift to G.729 once it is determined that they must traverse a WAN link?
Yes, this is very easy and common practice. Basically each phone is placed in a region. You define calls within that region (typically g.711) and calls between regions (g.729).
don't forget to rate posts that help you :)
I just wanted to clarify about this. I have a NM-HDV2-2T1 card in slot1 that we're planning to get a PVDM-64. Basically, that should be enough for 2 T1's and have 1 DSP left over for conferencing. Aren't we able to use the leftover for conferencing resources?