I use it PE side; with various customers on CE side.
Kudos to VoIPex on genius marketing; they've persuaded so many people that it's the be-all and end-off of VoIP, that we were forced - by customer demand - to deploy it and offer it as a service.
There are a few big reality checks that customers get:
(a) ViBE cannot fix bad bandwidth. This is the single biggest misconception that customers have and, despite the number of times I've explained this to them up front, I've still had some deploy it, cancel a month later, and then expect a refund on the equipment, because they've refused to listen to me after allowing someone else to persuade them that it can fix packet loss.
(b) ViBE is a trunking solution; not only does it have zero benefit if you don't generally make at least four simultaneous calls during routine usage, but it actually wastes a huge amount of bandwidth if your usage is often zero or one call at a time. Low-volume customers get horrified when they realise it's used up more bandwidth, not saved them any. It uses ~96kbit/s of bandwidth 24/7/365 even if you have zero calls, just on two-way monitoring, keepalives, etc. If you're going to deploy, ensure that it's at a site where there are sufficient simultaneous calls to warrant it. (e.g. >8 at peak times, >3 on average during office hours). Personally, I think it is a complete waste unless you need to exceed the simultaneous calls that the underlying link can physically carry.
There are also a few reality checks for providers:
(a) if you want to get the most out of ViBE, it needs to be used for both voice and data. That means that, unless you have your own IP-Connect (and I don't mean the use of someone else's branded as your own) you're going to tunnel your client's data over DSL bandwidth that you're paying for, into your network over capacity that you're paying for, and break out data back onto the internet that you're paying for. i.e. TRIPLE the cost of using a regular wholesale DSL account.
(b) Despite (a), it's ability to prioritise and split voice and data relies on consistency of performance of the underlying connection. E.g. If you assume your 4Mb/s ADSL will achieve 3.5Mb/s download and 0.4Mb/s upload, program the ViBE unit as such, and then discover that there is DSLAM congestion, the ViBE until will NOT be able to calculate the actual performance achieved over the DSL and dynamically reconfigure to reserve bandwidth for VoIP.
(c) The PE equipment, while pretty stable, is not flawless. In fairness, we typically get a number of months worth of uptime before suddenly having to reboot (and that has increased to around 6 months since reducing the load to a fraction of the hardware's stated capacity).
(d) When rolling out Hosted Switchboard and similar solutions, SIP signalling used for busy lamp's, presence, etc. causes short spikes of bandwidth usage that ViBE cannot reduce at all. It can only reduce bandwidth associated with call media, not signalling. Keep this in mind as it will impact on the expected benefits for Hosted Switchboard customers.
And reality for both:
(a) there was a time when an ADSL typically had a mere 256kb/s upload; ViBE was extremely useful to carry many calls on the same capacity. These days, 640kb/s or 1024kb/s upload speeds are typical. That's > 12 / >18 simultaneous calls. Reality check: clients spending that much money that they have > 12 simultaneous calls don't want to use ADSL. They typically have fibre connections or M.E. or carrier grade wireless. So ViBE isn't needed for smaller clients or larger ones any more.
(b) The hardware specs that VoIPex cite are way too generous. Literally halve the claimed simultaneous calls and remote licenses for each of the devices if you want them to work reliably, whether PE or CE equipment.
So, negatives aside, there are some places where ViBE is pretty useful:
(a) We provide service via many WISPs. ViBE allows us to build a tunnel right to the core of their networks and deploy a PUBLIC IP on the tunnel. That allows us to avoid NAT issues on lousy DSL routers and to perform bandwidth monitoring to that public IP (even if the ADSL connection changes or the WISP suddenly changes the backhaul connection, e.g. to 3G backup).
(b) By contrast, any other tunneling protocol (e.g. PPTP, L2TP) would typically double (or even triple) the bandwidth consumption.
(c) Those WISPs are often in rural areas where high-speed bandwidth is costly. Even those that use diginet or similar premium links pay a fortune on distance charges. Those that use ADSL for back haul often congest their local DSLAM. With decent bandwidth at a significant premium over what it would cost in the big cities, ViBE helps save money, reduce congestion and work within the confines of limited bandwidth.
So, with the good and bad together, I guess what I'm saying is that - provided you deploy ViBE where it makes sense (and not all over the show) - it is worth deploying. As broadband (ADSL, HSPA+ and LTE in particular) connection increase in capacity, there are fewer and fewer places where ViBE makes sense. But there are still instances where it makes sense and adds benefit. And, if you want a public IP at a customer site (e.g. for bandwidth monitoring, NAT avoidance, etc.), ViBE is the only way to tunnel VoIP without adding extreme overheads.
