If we have a telepresence unit connected to our MPLS provider (Telco A), and we're interested in having a conference with an outside company on a different provider (Telco B), what are some options to make this happen?
Is it possible to have dual telco providers (telco A&B) connected to our unit to allow the most flexibility?
Thanks for any info
Here is technical answer:-
Telco A customer (MPLS) can talk to Talco B (MPLS) customer. Provided Telco A & Telco B are connected through TelePresence Exchange. TelePresence Exchange solution is available. We have three TP Exchange POP present for global footprint and planning more.
If its businesses requirement:- Do let us know, we have the solution for same, if required we can talk to your existing Telco to connect it with our tp exchange.
Thanks for the info!
Who hosts the 'Telepresence Exchange' solution that you mention? For the 'business requirement' who hosts 'Our TP Exchange'?
We can discuss offline with more specific details. Please enable your contact info on mail to assist you further on "Telco A customer having TP session with Telco B customers".
I think you would have find the solution by now but was going through some telepresence documentation and you can ask your service proviver if they offer Business to Business Telepresence services.Service provider have SBC (Session Border Controller ) and you need CUBE (Cisco Unified Border Element) .
You can check with Cisco how much it is going to cost you.
Go through this link may be useful in future when you are looking again. Some part of that document.
"The growing list of service providers that have announced intercompany Cisco TelePresence service include AT&T, BT, Orange, Tata, Telefonica, and Telstra."
(because I don't know the details about the Cisco standard proposal solution [TelePresence Exchange]
for this particular situation mentioned earlier) ...
if needed, you can use two or more suppliers
(for example, MPLS providers or Providers supply other technologies such as ATM / FR)
you will have to guide the traffic of your solution TP until the last device on your network
facing yours MPLS Providers (two or more)
if you have two suppliers, you can use two different interfaces (one for any targeted Provider)
as a less expensive option,
or use two Core devices (physically separated, but logically attached by a tunnel for example or
any other redundant Cisco IOS software solution, to provide redundancy)
(if you don't provide redundancy, this could be dangerous because you are generating
a single point of failure in your environment)
then, manipulate BGP attributes to "mark" the traffic TP (routing level)
so that you can redirect this traffic toward one or another provider,
depending on the location the destination of such TP traffic (destination sites over other providers)
so you can see sites located in two or more different destination provider
(probably this approach is not so scalable, but if we have the appropiate knowledge of the network
could be practical with a regular number of sites)
If you choose this kind of deployment you would need to "convert" your network equipments [whe use two, facing 2 differents providers]
to become a "peering point" BGP (to share the routes / routing information; very likely that/those device(s) already working in this way)
well, you can use this "peering point" BGP device as the "traffic controller"
that "BGP speaker" will choose the right provider and "place" the traffic out the appropriate interface
(these "interfaces" [link] need to meet the requirements required by TP: bandwidth, delay, jitter, etc)
when the TP traffic from the site located over another different provider ( to our provider which supports
our source TP site), recibe the back TP traffic, it will revert the process and TP traffic from the external
destination TP site will be redirected to the appropriate source TP (behind our site in our original provider)
This solution works in a similar way to a process of route redistribution
(and therefore the Network Admin's Team must have a good understanding of the process of redistribution
of routes BGP and IGP protocols, and know the traffic behavior)
another very important aspect to consider is that these solutions require that TP traffic
must be labeled with the best quality of service available
(generally a defined POLICY with many CLASS, as follow
CS5 QoS for voice / video - QoS CS4 and CS3 for QoS signaling)
keep in mind that providers can ignore these Client QoS defined classes
and they (the Providers) could use their own "marks"/ "class" to private/internal use
( over the MPLS, ATM / FR, etc. clouds ) so these provider's "marks"/"classes" could be
with a lower quality service.
therefore, is necessary to define an explicit SLA
to provide/meet specifics TP Client Traffic requeriments / needs
(to avoid un-appropiated marking/classing, from us to the provider cloud,
and over the provider cloud , too)
Remember, QoS Policy/Class/Mark need to be define and apply over the appropiated interfaces
(and over the appropiated places over our internal infrastructure)
to do it right, will be necessary have a correct understanding of TP traffic behavior
(and the other kind of Traffic) within our own network infrastructure
I hope that you find it useful