Campus Design

typesh

Hey


I am trying to set up a topology to play with and trying to stick to the Campus LAN.

I am using Packet Tracer and have (4) 2960 along the bottom of the screen. These are my Access Layer switches. Right above those 4 switches I have (2) 3560 switches set as my Distribution Layer switches. Above the Distribution Layer switches I have 1 router. I am guessing this Router is my Core layer? This entire network is using 192.168.1.0/24 and I am going to configure the default gateway to be 192.168.1.1

I am confused here: Since I have 2 Distribution Layer switches, then each switch needs its own connection to the router. One router interface can be 192.168.1.1 on Fa0/1, but the other 3560 will have to connect to the router on Fa0/2. This is because the Router's Fa0/1 has already taken 192.168.1.1. This will mean that the Router's Fa0/2 will have something like 192.168.1.2 as the interface address, yet the PCs are configured to use 192.168.1.1 as the default gateway. Suppose the link from the 3560 to the router's Fa0/1 goes down. The PCs will still send packets addressed to the default gateway instead of choosing to use 192.168.1.2

How do we get around this?

The router cannot use 192.168.1.1 on Fa0/1 and on Fa0/2 as there is an overlap, so I am not sure how to solve this.

thenjduke

Where are you trying to route too? This is all in the same subnet. Think about the OSI Model. You are not routing anywhere. Now if you needed to route you have another interface that was going out to your ISP or MAN.

thenjduke

I am actually glad you posted this.. Made me think a little bit as I am getting ready for the test.

typesh

Sorry I should have also stated where I am trying to route to.

Imagine an identical topology right next to this one. This second topology will be 192.168.2.0/24

I will also be adding a 3rd identical topology beside these two. This third topology will be 192.168.3.0/24

The WAN link between topology 1 and topology 2 will be 10.1.1.0/30

The WAN link between topology 2 and topology 3 will be 10.1.1.4/30

thenjduke

Then you would have one fast ethernet being the 192.168.1.1 and then the other interface that you are connecting to the 192.168.2.0 would be 192.168.2.1 and then you would setup the routing protocols and the paths. The default gateway on the 192.168.1.0/24 network would be 192.168.1.1 and the router would know that you are looking to route to 192.168.2.0 network and so on with the route paths you setup or they learn.

thenjduke

Also if you are connecting both switches to the fast ethernet interfaces they your default gateway would be ther other interface you are routing out of.

Router
FA 0/0 FA 1/0 Serial 0/1 (The serial would be default gateway)

jscimeca715

You would need a core switch above the two distribution switches. To my knowledge, there is no such thing as a Core router when it comes to ICND 1 or 2. You can create a secondary address under the routers fastethernet interface and use that but it would still need to be in a seperate subnet.

thenjduke

also one more thing you will overlap on the interfaces as well.

typesh

Here is what I have going on so far...

thenjduke

Take a look at this document. http://www.cisco.com/warp/public/cc/so/neso/lnso/cpso/gcnd_wp.pdf
Will give you some more insight. You will not be able to connect both Distrubution Switches to the router as if you try to give the interfaces on the router the ip address in same subnet the ip address will overlap. I would put a layer 3 switch above the distro switches for the core and do redunant links to the distro switches.

networker050184

Since you are already using 3560's why not set them up as the default gateways for the subnet? Might be a little over your head, but use HSRP between the two on SVIs and then you can use L3 uplinks into your core router.

mikej412

typesh wrote: »

Suppose the link from the 3560 to the router's Fa0/1 goes down. The PCs will still send packets addressed to the default gateway instead of choosing to use 192.168.1.2

How do we get around this?

You find a CCNP to implement redundancy for you using HSRP, VRRP, or GLBP -- but it may not be supported in Packet Tracer.

jbrad95706

thenjduke wrote: »

Take a look at this document. http://www.cisco.com/warp/public/cc/so/neso/lnso/cpso/gcnd_wp.pdf
Will give you some more insight. You will not be able to connect both Distrubution Switches to the router as if you try to give the interfaces on the router the ip address in same subnet the ip address will overlap. I would put a layer 3 switch above the distro switches for the core and do redunant links to the distro switches.

Good stuff in that link.

typesh

Hmm.. Don't really understand HRSP and SVIs.. Still working on ICND1 material.
I think what I will do is get rid of the 1 distribution switch and forget about redundancy at the distribution layer for this lab.
Thanks for the article as well.

mikej412

typesh wrote: »

I think what I will do is get rid of the 1 distribution switch and forget about redundancy at the distribution layer for this lab.

You may not know how to configure redundancy to remove one router or router interface as a single point of failure -- but you should be learning about switching and spanning tree. Taking out 1 distribution switch removes some layer 2 redundancy.