Automatically Switching a Tunnel to the Primary Interface
When two network connections (primary and backup) are used, a tunnel automatically switches to the backup interface if the primary connection fails.
Once the primary connection is restored, however, the tunnel continues to use the backup interface because the existing tunnel session remains active and is not automatically re-established.
On Donyx routers, this behavior can be addressed by using the Pinger service together with the Events mechanism to re-establish the tunnel when the primary interface becomes reachable again.
| The configuration can be performed either through the dnxOS web interface or the CLI. For clarity, this article uses the CLI. |
Example
The following example uses this network configuration:
-
port3 — primary wired interface;
-
wifi1 — backup wireless interface.
Under normal conditions, the tunnel should operate through port3.
If connectivity through port3 is lost, the tunnel should automatically switch to wifi1.
Once port3 becomes available again, the tunnel should automatically switch back to the primary interface.
Step 1. Configure interface priorities
Assign different metric values to the interfaces. The interface with the lower metric has the higher priority.
/ip interface port3 metric 203
/ip interface wifi1 metric 206
Step 2. Create a connectivity check
Create a Pinger service to monitor the availability of the remote L2TP server through the primary interface.
/service pinger add interface=port3
host <l2tp_server>
apply
Step 3. Create an event
Create an event that is triggered when the primary interface becomes reachable again.
When the Pinger service changes to the success state, the L2TP tunnel is re-established.
/event add trigger=pinger
target port3
value success
action /tunnel l2tp <tunnel_name> apply
apply
After the tunnel is re-established, routing is recalculated based on the configured interface metrics. Since port3 has the lower metric, the tunnel is established through the primary wired interface.
Result
After completing the configuration, the tunnel operates as follows:
-
The tunnel operates through port3.
-
If port3 fails, the tunnel automatically switches to wifi1.
-
When port3 becomes reachable again, the Pinger service detects the status change.
-
The event re-establishes the tunnel.
-
The new tunnel is established through port3 because it has the higher priority.
Recommendations
| The event is triggered only once when the Pinger service changes to the success state. |
To avoid false or premature triggers, configure the Pinger polling interval and retry count so that its state changes only after the tunnel has been successfully re-established through the backup interface. This helps prevent unnecessary tunnel re-establishment during brief network interruptions.