A down-recovery timer can be set so that if the cellular MDA fails to establish cellular service or if BGP cannot be established on any interface configured in the system, the node will reboot. The down-recovery-interval is configured at the cellular MDA level and is not specific to a particular SIM. The timer can be set when there is a single SIM or two SIMs installed in the node.
The operator can specify the criteria that will cause the node to reboot by configuring the down-recovery-criteria command. The down-recovery-criteria can be set to port or bgp. When set to port, all cellular ports configured on the system are monitored during the down-recovery interval. When set to bgp, all BGP sessions configured on the node are monitored during the down-recovery interval. Both options can be specified concurrently, and the node will use either the cellular port state or BGP session state to declare the operational state of the node as down.
When set, the down-recovery-interval specifies the length of time that the configured down-recovery-criteria are monitored from the moment when either the ports or the BGP sessions are declared operationally down. If the interval is exceeded without any port or BGP sessions going operationally up, the node reboots so that the preferred SIM can try to connect to a cellular network again. As soon as any port or BGP sessions is operationally up, the down-recovery timer stops.
The down-recovery-interval is measured in minutes, with a range of 1 to 240 minutes. Sixty seconds before the timer expires, the node will issue a log event stating that the node will reboot in 60 seconds if the down-recovery condition (based on the configured criteria) is not resolved. This 60-second warning interval can be used for further debugging and diagnostics before the node resets.