Convergence: measuring what your network actually does
Everyone claims sub-second failover. Very few have measured it.
Writing
OSPF, BGP, static and policy routing, VRFs and multicast: how routes are chosen, how they fail, and how to read what a router is actually doing.
Everyone claims sub-second failover. Very few have measured it.
A default route means "send it here and hope". Sometimes hope is misplaced.
One router, several routing tables that cannot see each other. Where that helps and what it costs.
Dynamic routing is not automatically better. Sometimes the static route is the professional choice.
It is not a metric and it never leaves the router. That is the whole idea.
Two protocols feeding each other at two points will create a loop. Here is how to stop it.
You may not need BGP. If you do, the parts that matter are few.
The adjacency state tells you which check failed. Work the list in order.
Traffic prefers a slow link over a fast one. The cause is usually a default reference bandwidth.
Areas are not organisational tidiness. They control how much work every router has to do.