A provider is down
A provider is the carrier calls go out through (or come in from). When one has trouble, calls that route through it start failing — but a well-set-up routing plan (LCR) often moves on to a backup on its own, which changes what "down" even looks like to a customer.
The providers list
Open Routing > Providers for an overview of all of them at once.
Two columns matter here, and they answer different questions:
| Column | Meaning |
|---|---|
Register | Live outbound registration state for providers that need it: Registered (healthy), Rejected (the provider refused it — a real problem), or a dash if this provider doesn't register at all (common for wholesale/IP-authenticated providers, which is not a fault). |
Status | Whether this provider is administratively active or disabled here in the system. A provider can show active and still be unreachable on the network — this column doesn't test connectivity. |
A provider's own registration card
Open the provider (only ones with outbound registration switched on show this card):
This is the same Registration status card used on devices (see A phone does not register) — Online · Registered means the provider connection itself is healthy right now.
How LCR failover hides a down provider
A routing plan (LCR) lists providers in order and can carry a separate Failover provider, used only once every regular one in the list has failed. Open Routing > LCR / routing and the plan in question to see the order and whether failover is switched on.
| Field | Meaning |
|---|---|
Order | The strategy this plan uses to try its providers — see LCR & routing rules for the four options. |
| Provider table | Every provider in this plan, in the order they're tried, each individually Disable-able without removing it from the list. |
— no failover — checkbox | When checked, only the first provider in the row is ever used — if it fails, the call simply fails rather than trying anyone else. |
Failover provider | One extra provider tried only after every regular provider in the list above has failed. |
Because of this, a provider being down doesn't always mean customers notice — if a working backup is next in line (or set as the failover provider), calls keep connecting through it instead, just possibly at a different rate. That's exactly what the Attempts (cause routing) trace on a call's detail page shows happening: see A call fails.
If no working provider is left at all for a call, its hangup cause reads 204 · No suitable providers found — see the code tables in A call fails.
Check
Open the provider and confirm its Register state (if it uses registration), then open one of the affected calls and read its Attempts trace — it shows exactly which provider was tried, what it returned, and whether the system failed over to another one on its own.