Cisco's Linux version of the Call Manager is released (Version 6.0), but
it is considered in the "Early Adopter" phase right now.
Avi Baumstein wrote:
> On Tue, 2007-09-11 at 16:21 -0400, Brian Bartholomew wrote:
>> If you have a building full of the UF-recommended brand of VOIP phones
>> on the same LAN, can they talk to each other without involving any
>> additional central points of failure? I'm not asking for conference
>> calls, voicemail, recordings with beeps, queues, forwarding, or
>> whatever. Just using the last quad of the IP as an extension number.
> The server piece in a Cisco IP phone solution is a 'Call Manager' which
> would be the central point of failure you refer to. After the Call
> Manager sets up a call (looks up the number you dialed, cross refs with
> it's IP, rings the phone, then passes it off), the actual voice
> connection is directly between the two phones (if within the same phone
> system) or between your phone and the gateway if one end crosses the
> telco. If the Call Manager takes a dive during a call, the call should
> not be interrupted.
> The bad news (since this is the Linux list :) is that the Call Manager
> runs on Windows (in the current version, next version is Linux). The
> good news is that Cisco recommends (and I understand CNS has
> implemented) multiple redundant Call Managers, so that the system can
> tolerate failures of individual pieces.
> I'm also fairly certain that the CNS VoIP implementation is using SCCP
> ('Cisco Skinny'), so the SIP dial-by-IP functionality isn't present.
> All the same goes for the Health Science Center VoIP system, of which
> I'm much more familiar. I will say that in the 6 or so years we've been
> running the Cisco phone system in the HSC, it has been extremely
> reliable, give or take a few times Bellsouth has screwed up a PRI (which
> actually isn't much of a problem as long as there is capacity on
> non-screwed-up PRIs). Yes, parts have died, but service outages have
> been prevented by the built in redundancy.