Users Guide

A tracked route matches a route in the routing table only if the exact address and prefix length match an entry in the routing table. For
example, when configured as a tracked route, 10.0.0.0/24 does not match the routing table entry 10.0.0.0/8. If no route-table entry has
the exact address and prefix length, the tracked route is considered to be DOWN.
In addition to the entry of a route in the routing table, you can configure how the status of a route is tracked in either the following ways:
• By the reachability of the route's next-hop router.
• By comparing the UP or DOWN threshold for a route’s metric with current entries in the route table.
Track Route Reachability
If you configure the reachability of an IP route entry as a tracked object, the UP/DOWN state of the route is determined by the entry of
the next-hop address in the ARP cache.
A tracked route is considered to be reachable if there is an address resolution protocol (ARP) cache entry for the route's next-hop
address. If the next-hop address in the ARP cache ages out for a route tracked for its reachability, an attempt is made to regenerate the
ARP cache entry to see if the next-hop address appears before considering the route DOWN.
Track a Metric Threshold
If you configure a metric threshold to track a route, the UP/DOWN state of the tracked route is determined by the current metric for the
route entered in the routing table.
To provide a common tracking interface for different clients, route metrics are scaled in the range from 0 to 255, where 0 is connected
and 255 is inaccessible. The scaled metric value communicated to a client always considers a lower value to have priority over a higher
value. The resulting scaled value is compared against the threshold values to determine the state of a tracked route as follows:
• If the scaled metric for a route entry is less than or equal to the UP threshold, the state of a route is UP.
• If the scaled metric for a route is greater than or equal to the DOWN threshold or the route is not entered in the routing table, the
state of a route is DOWN.
The UP and DOWN thresholds are user-configurable for each tracked route. The default UP threshold is 254; the default DOWN
threshold is 255. The notification of a change in the state of a tracked object is sent when a metric value crosses a configured threshold.
The tracking process uses a protocol-specific resolution value to convert the actual metric in the routing table to a scaled metric in the
range from 0 to 255. The resolution value is user-configurable and calculates the scaled metric by dividing a route's cost by the resolution
value set for the route type:
• For intermediate system to intermediate system (ISIS), you can set the resolution in the range from 1 to 1000, where the default is 10.
• For OSPF, you can set the resolution in the range from 1 to 1592, where the default is 1.
• The resolution value used to map static routes is not configurable. By default, Dell Networking OS assigns a metric of 0 to static
routes.
• The resolution value used to map router information protocol (RIP) routes is not configurable. The RIP hop-count is automatically
multiplied by 16 to scale it; a RIP metric of 16 (unreachable) scales to 256, which considers the route to be DOWN. For example, to
configure object tracking for a RIP route to be considered UP only if the RIP hop count is less than or equal to 4, you would configure
the UP threshold to be 64 (4 x 16) and the DOWN threshold to be 65.
Set Tracking Delays
You can configure an optional UP and/or DOWN timer for each tracked object to set the time delay before a change in the state of a
tracked object is communicated to clients. The configured time delay starts when the state changes from UP to DOWN or the opposite
way.
If the state of an object changes back to its former UP/DOWN state before the timer expires, the timer is cancelled and the client is not
notified. If the timer expires and an object’s state has changed, a notification is sent to the client. For example, if the DOWN timer is
running when an interface goes down and comes back up, the DOWN timer is cancelled and the client is not notified of the event.
If you do not configure a delay, a notification is sent when a change in the state of a tracked object is detected. The time delay in
communicating a state change is specified in seconds.
Object Tracking
511