We're planning on spliting our UC environment (CUCM, UCXN, UCCX) across a WAN. By looking at the SRND's for each, I estimate that we'll need 3.7Mbit bandwidth on the WAN just for intercluster communication for all the servers. We have a 6Mbit connection, so that leaves us 2.3Mbit for actual voice traffic if we were to terminate calls on the other side of the WAN from the users.
The region settings between the WAN-connected sites are set to use G.729 for all calls, except for UCCX (this requires G.711, so it has it's own region). My question is, when the call is sent from UCCX to the Agent, will this downgrade the call quality to G.729? Meaning, the only time the calls will be using higher bandwidth is when they are in queue with UCCX (G.711), and once the agent takes them, they will reduce bandwidth to G.729.
We have 2-node clusters of CUCM, UCXN, and UCCX. 1 node of each cluster will be in each physical site, separated by 6Mbit WAN (MPLS).
Are you saying we need 1.5Mbit for each node (total of 3Mbit), or between the nodes (1.5Mbit total) for the CUCM and another 1.5Mbit for ICCS? This was a little hard to grasp in the SRND as well, which is why I came here.
So basically we have:
Unity currently has 32 ports configured. UCCX typically has 5-20 calls on hold during peak, and agents toping out at about 75 peak concurrently logged in.
I'm getting the sinking feeling that 6Mbit isn't going to cut it, but I don't have a solid tool (or understanding of how to) profile the current ICCS traffic footprint. I can easily calculate bandwidth needed for voice traffic, but the ICCS stuff isn't very deeply described in the document.
I'm not able to access my old voice mail messages all of a sudden. The recording says something like 'the message is currently not available'. This has never happened before in all the years I have been using this system. I have t...