Symptoms: Understanding IP On-The-Fly vs. IP Standard modes

Facts:

• NBX 100
• SuperStack 3 NBX
• Remote IP Phone
• NBX Phones
• License
• License Key
• IP On-The-Fly
• Call Setup

Fixes: Understanding IP On-The-Fly:

IP On-The-Fly for NBX is the ability of NBX phones/devices (that are on the same subnet as the NCP, and idle as a Layer 2 device) to switch to IP mode during call setup, when involved in a call with any device that is on a remote subnet (a Layer 3 device). Because the NCP drops out of the call completely after call setup, and both phones/devices communicate directly to each other, both phones must talk either at Layer 2 (MAC to MAC) or Layer 3 (IP).

Simply put, if a Layer 2 phone is called or calls a remote IP phone, it must act as a layer 3 devices in order to talk for the duration of that call, so the NCP will temporarily (On-The-Fly) assign an IP address so the phones can talk IP to IP. This is only required when:

1. a conversation between a local phone/device and a remote IP device initiates

2. a local phone is conferenced with at least one remote IP device

The IP On-The-Fly address pool is populated in NetSet, and these addresses therefore are ONLY used to assign devices on the same subnet as the NCP (devices normally idle at Layer 2). IP OTF pool addresses are never given to the remote IP devices, which always maintains its own IP address, configured one of three ways:

1. Through LUI (programmed directly into the Local User Interface of the phone)

2. Via DHCP server serving the remote site devices

3. Manually assigned in NetSet with an IP appropriate for that remote subnet

Consider this as opposed to IP standard. Suppose there are 100 local phones, and two remote sites each with 3 IP phones. Those phones maintain a constant IP address. If you implement IP standard, you will have to manually assign 100 IP addresses, one each for every local devices, “just in case” it needs to be in a call and communicate with a remote IP phone/device. For most calls (being calls between extensions at the main site), phones talk Layer 2 (despite the fact they may have an IP address assigned), and the IP address is not used.

The NCP knows that both phones/devices are local, and will set up each device to talk MAC to MAC (L2). Only if a local device is on a call with a remote IP device, the IP address is used. So you need 100 IP addresses pre-configured for all local devices. With IP On-The-Fly, you can theoretically have only 6 IP addresses in the pool, as at most there would only be 6 conversations at any one time involving a remote IP device (as there are only 6 remote IP phones). So the efficiency goes way up, as does the administration of IP addresses.

An IP address from the pool is assigned On-The-Fly to a local phone/device when it is on a call with a remote IP device for the duration of the call only. Meaning the local phone will switch to IP mode and send it's audio packets directly to the remote IP device during the call, and when the call terminates, that IP address is recycled to the pool, and the local phone returns to Ethernet binding.

This can be seen “live” by bringing up the Device Configuration, IP settings, of a local phone during a phone call in progress with a remote IP device. You will see the Binding change from Ethernet to IP, and you will see an IP address populate the field “IP Address Reported By Device” showing the IP OTF address assignment for that call. Notice when the call is terminated, the binding changes to Ethernet, and the IP address is de-assigned.

Remember that the NCP is involved only in call setup/teardown; after the callers connection, the devices talk directly to each other, the NCP can theoretically be pulled from the chassis and the call will maintain. But every action of the phone, be it raise the volume, switch to speakerphone, or even hang up, requires the NCP.

NOTE: The NBX is efficient! If two phones are on a remote IP subnet, meaning they have statically assigned full time IP addresses at that remote site, and they call each other, the NCP knows (based on the “IP address Reported by Device” information) that these devices are on the same subnet, albeit a remote subnet, and actually tells the remote IP phones to “downshift” and talk MAC to MAC Layer 2, simply because they can and it is more efficient. If one remote phone calls a remote phone at a different remote site, they NCP knows they are not at the same site (subnet) and they must maintain their IP (L3) communication during that call.

• Product(s): NBX
• Sub Product(s): NBX 100, NBX Phones, SuperStack 3 NBX

— David Gonzalez 2021/04/16 10:26