I do not know the details of her problem. The routers in networks often need to be considered Single Point of Failures (SPoFs).
For this reason, the concept of the "Virtual Router" was developed. In this regard, we speak of VRRP (Virtual Router Redundancy Protocol); see for this RFC 5798.
According to the VRRP concept, a SPoF can be set up redundantly - namely:
either according to the Master-Slave Principle or according to the Load Balancing Principle.
In my chapter on VRRP, the two principles are illustrated:
Fig. 004484: Master-Slave Principle (only one router is active; cold reserve principle) .
Fig. 004485: Load Balancing Principle (both routers are active)
I am trying to propose a caching strategy for content-centric networks. where caching will take place on edge of the network. problem arises when an edge node fails the whole network crashes. i dont think concept of virtual Routers can be applied here.
that was just an idea. According to the concept of "virtual router", a "virtual proxy" can be set up, namely: router = proxy of caches. Several cache systems can be connected to such a virtual proxy.