Hi,
so our brand new and shiny BE6KS just arrived, and I'm going to set it up the next couple of days. I already have some lab experience with a CUCM - however in our production environment we are migrating from a UC560.
I went through the available documents, especially the Midmarket BE6000 CVD, the PreConfiguration Guide and the relevant parts of the BE6KSIG setup guides.
What I was missing is a basic overview on how the stock configured BE6KS should be integrated on a LAN. However Andrew Hickman posted some real useful information about this here: https://supportforums.cisco.com/discussion/12450581/typo-be6000s-quick-start-guide
Ciscos take on the appliance is to have it as a small self contained blackbox with a tiny 172.27.199.0/26 subnet. CIMC, ESXi, ISR all listen on this subnet and have the default route out of GE0/0 of the ISR. GE0/1 is, what I would call it an access/maintenance port if you need direct access to this tiny subnet for setup reasons. With this setup I see the BE6KS as a "leaf" in my network - what I personally don't like that much. Especially what I don't like too much about this setup is, that the voice VLAN for your IP-Phones has to go to through a central router to be able to talk to the UC Apps. This might be a bottleneck. Also managment of the ESXi (and possibly VM-Backups) all go out through the shared GE0/0.
So my take on the Setup would be the following (we have DataVLAN1, GuestVLAN2, VoiceVLAN100) :
UCSE-MGMT: Use this port for CIMC and have it connected to our DataVLAN (there is no way to have CIMC bridged to UCSE-GE2 right?)
UCSE-GE2: Use this port in our normal DataVLAN, and have the ESXi-Mgmt connected to this port.
UCSE-GE3: Use this port for the UC-Apps and have it connected to the VoiceVLAN
GE0/0: Use this port for the ISR (config, MGCP) have it connected to the VoiceVLAN
A router, which is already present will route between Data and VoiceVLAN so UC-Configuration, Webservices etc. are reachable from all of our data VLAN.
Yes, that's 4 lan-cables coming out of the box than rather one, but I like to have my things physically structured.
The main advantage is, that UC-Apps and Phones are on the same subnet and on the same vlan. So if our router has performance issues or dies or just has to be rebooted telephony service won't be impaired.
What do you think? Is this a valid approach for a small scale deployment? What could be improved?
Regards,
Fabian
