@Mark Malone makes an interesting suggestion. But it is radically different from the environment presented by the original poster, especially his addition of 2 routers in a question that was asking about how to implement a second core switch. If you have the resources to implement it then you would have an excellent solution.
@balaji.bandi presented a drawing that has the important aspects of implementing a second core switch to provide some redundancy for the network. Those aspects include:
- a second switch for core that has capabilities similar to the first core switch. (same model of switch is the optimum solution but a switch model with similar capability will work)
- an EtherChannel connecting the core switches. This means that each switch has at least 2 interfaces in the EtherChannel which provides capacity and most importantly failover if there is a failure of a single switch interface.
- a connection from the new core switch to the set of downstream switches. This would be configured as a trunk and would carry all of the vlans defined in this network.
The suggestion from BB does not address some aspects of this redundancy, such as what will be the default gateway for the downstream clients in each of the vlans? HSRP would typically provide this part of the solution.
BB provided the main points of a solution and suggested that the original poster draft a config which we could then make comments for improvement. The original poster seems reluctant to do that. The config of the new core could be very similar to the config of the original core, the difference being allocating 2 interfaces for the EtherChannel, configuring EtherChannel on the 2 interfaces (and perhaps introducing HSRP on the vlan interfaces).
HTH
Rick
HTH
Rick
