Telecoms glossary • F
Failover routing
The rule that decides where a call goes when the office cannot answer it. It runs upstream of your building, which is the only reason it still works when your building is dark.
What is failover routing?in VoIP and business telephony
1. Failover routing is a pre-set rule that sends an incoming call somewhere else when the intended destination cannot take it. On a hosted phone system the rule runs on the provider’s platform rather than on your premises, so calls are still answered when your connection, your power or your handsets are out.
2. That location is the whole point. A rule held on equipment in your comms cupboard cannot redirect anything once the cupboard loses power, which is why an on-premise system needs the same job done upstream by the carrier.
How a failover rule fires
Failover is not a mode the system switches into. It is a list of conditions evaluated on every inbound call, and the call falls down the list until something answers.
-
The call reaches the platform
Your provider’s system receives the call before your site does. Everything that follows happens there, whether or not your office is reachable.
-
The destination is tested
The platform tries the mapped extension, queue or hunt group. The usual trigger is that the device is unregistered, meaning it has stopped checking in over SIP. Busy and no-answer timers are separate conditions with separate rules.
-
The next rule applies
Unreachable sends the call to the first fallback. That is typically a mobile number, another branch, or a queue on a site that is still up.
-
There is always a final destination
If nothing answers, the call lands somewhere deliberate: voicemail to email, an answering service, or a recorded message with an alternative number. A rule chain that ends in ringing is a rule chain that has not been finished.
How quickly the platform notices depends on the registration interval configured on your extensions. A device that re-registers every few minutes is marked unreachable within minutes, not seconds, so the first call or two after an outage can still ring into nothing.
Building the layers
Failover routing is one control in a chain. On its own it moves calls off a dead site; combined with the layers below it can keep the site alive instead.
- Keep the network up
- A UPS on the router, the switch and the access points holds the site through a short interruption. Handsets drawing PoE from a protected switch stay registered, and failover never has to fire. Size it on the actual load and the outage duration you expect, not on a shelf label.
- Keep the connection up
- A second link on a different medium and a different operator, usually LTE or 5G behind fibre. The router fails over, the handsets re-register on the backup path, and calls continue at reduced quality. A backup from the same operator over the same route is not a backup.
- Move the people
- The provider’s mobile app or a softphone on a staff phone keeps the extension alive on mobile data with nothing at the office running. For many businesses this is the whole disaster plan and it costs nothing extra.
- Move the calls
- Divert to mobile numbers, to another branch, or to an outsourced answering service. This is the layer that works when everything else has failed, and it is a configuration option rather than a product.
Where the platform itself is hosted matters too, which is what geo-redundancy covers. Failover routing protects you from your own outage; geo-redundancy protects you from theirs.
What actually takes a South African site off the air
The reasons a business needs this have shifted, and buying against the wrong one produces the wrong plan.
- Infrastructure crime
- The largest and fastest growing cause. ICASA’s 2025 State of the ICT Sector report put theft-related costs to South African operators at R201.48 million, up 189% from R69.59 million the year before. A cut fibre run takes a whole street off, and no provider can route around a hole in the ground quickly.
- Power interruption
- Eskom reported 441 consecutive days without national load shedding as at 4 August 2026, so this is no longer the daily planning assumption it was. Municipal load reduction, local faults and substation failures continue, and they hit every operator and every business in the affected area equally. Plan for hours without power, not for a schedule.
- Single-supplier dependency
- One fibre line, one operator, one route. The failure does not have to be dramatic to be total.
- Ordinary equipment failure
- A dead router, a failed switch, a firewall change made on a Friday. The most common outage of all, and the one people never write a plan for.
Test it, do not assume it
A failover rule is invisible until the day it matters, and a rule that was configured once at installation has usually never been fired. Unplug the router, phone your own main number from a mobile, and time how long it takes before something answers. Then repeat it after any change to staff, mobile numbers or branches.
Ask a provider two questions at quoting stage. First, how long after a site goes unreachable does the platform divert, and is that interval something you can shorten. Second, what happens to a call that is already in progress when the line drops, because failover handles new calls and almost never rescues live ones.
About this entry
Definitions are written for South African business buyers and checked against primary sources – provider documentation, standards bodies and ICASA – not vendor marketing. Reviews are independent and sponsors are always disclosed. Read our editorial policy and scoring methodology.
Spotted something out of date, or think we have this wrong? Tell the editors and we will re-check it.