Cisco Campus Network Design Basics

Hello Brian

It really sounds like you’re a hard core networking guy. That’s great! You know, I’d say that Cisco engineers in general share a unique camaraderie that’s unlike anything else. Others can’t understand why we like networking so much! It’s just cool! I understand you completely.

As for the opposition you face when suggesting best practices in networking, it’s usually the case where you’d have the administration opposing changes “since everything already seems to work.” It’s when things fail because of lack of redundancy, or voice and data on the same VLAN or other similar issues that they come back and say “well why didn’t we do it right the first time?” It’s a matter of making them understand the issue in simple terms, which is not always easy! :slight_smile:

I’m glad we can be of help to you here at Networklessons. This really is a great place for network-loving engineers to dig deeper into the things we like so much…

Keep at it and looking forward to continuing our discussions!

Laz

1 Like

Hello Laz,
I have a OSPF design question mainly and I am going to use the below topology as the reference.
image

In this design, I have a pair of Campus core and a pair of data center core routers. Here I am running OSPF area 0 between Campus core and Data center core devices. Also I am using different areas between Core switches and different Distribution blocks. I am configuring Totally NSSA in all the areas other than area 0 so every area will have only the default route from the core switches. Core routers will advertise each other only the summary routes of different distribution blocks. For instance, Campus core pairs will advertise summary of different distribution blocks to the Data center core routers and vice versa. Therefore, every distribution switch will have only one route in their routing table that is only the default route and core routers will advertise only the summary routes of different distribution blocks to each other. What do you think about this design? Do you think I should modify anything or it is one of the best practices?

Thank you so much in advance.

Best Regards,
Azm

Hi Azm,

Using totally NSSA on your distribution layer is fine, these devices only need a default route to get anywhere else. You need to take a closer look at your core layer though…

There’s the “build triangles, not squares” saying. For example, your data center core1 router will always use campus core 1 to get to distribution layer 1 or 3. If you have an additional link from data center core 1 to campus core 2 then it can use both campus core 1 + 2 and load balance traffic. The same thing applies for data center core 2 to campus core 1.

Rene

1 Like

Thanks a lot Rene…

Azm

I just went through the lesson. I don’t understand the part which says that - “layer 2 links should be configured between the distribution layer”. The reasons are given. But i don’t understand that part. Could you please help me with it ?

Hi Sriguruprassad,

Let’s look at just one example why you might want to use L2 between the distribution layer switches. Consider this design:

https://networklessons.com/wp-content/uploads/2015/03/distribution-l3-access-l2.png

In this design, we have VLAN 10 on both access layer switches. All links on the distribution and access layer are layer two links.

Think for a minute about spanning-tree…let’s say that the left distribution switch is the root bridge. What will be the root port on all of our switches? Something like this:

campus-l2-distribution-layer-root-ports

Now imagine we have an L3 link in between the distribution layer switches:

campus-l2-distribution-layer-root-ports-wrong

Since the link in between the distribution layer switches is now L3, the distribution layer switch on the right side will select another interface as the root port for VLAN 10.

Now imagine you configure something like VRRP…this means that all VRRP traffic from the right distribution layer switch goes through the access layer to get to the left distribution layer switch. Not a good idea :smile:

Hope this helps!

Rene

2 Likes

Hi Rene,
i was struggling with the parts:

If one of the uplinks from the access to the distribution layer fails, VLAN 10 could become isolated.

The switches on the distribution layer will use a protocol to create a virtual gateway IP address. We need layer two connectivity for this.

If the link from Access switch 1 (left) towards Distribution Switch 1 (left) fails, then the link Access Switch 1 -> Distribution switch 2 is used. If there is no L2 link, the link Distribution Switch 2 -> Access Switch 2 will become the root port and traffic is forwarded to the Distribution Switch 1 (HSRP active node). This is not a good design, but the Vlan 10 would not be isolated.

Please correct me if i’m wrong.

Many thanks,
Oliver

Hello Oliver,

You are correct, VLAN 10 won’t be isolated since it can still go through the distribution layer switches. Here’s a picture to visualize this in case anyone wonders what this is about:

root-port-through-access-layer

In the picture above, there is no L2 link between the distribution layer switches and one uplink from the access layer to distribution layer switches failed. If the left distribution layer switch is the root then traffic will go through the access layer like this.

I just removed the sentence about isolation, thanks for sharing!

Rene

Hello Rene,

thanks a lot for clarifying. :slight_smile:

Cheers,
Oliver

what is the backplane in the specification and the word require high bandwidth and throughput, what does exactly mean

Hello Pipat

The backplane is the name given to the internal circuitry of a switch which acts as the pathway between individual ports. If you have a switch with 48 1Gbps ports, and you have computers connected to all of those ports, theoretically, you should be able to have 24 computers each sending 1 Gbps of information to the other 24 computers simultaneously. This requires an internal bandwidth on the backplane of the switch of 24 Gbps in each direction, for a total of 48 Gbps. If the switch has additional uplink ports, those must also be taken into account as they can add traffic to the backplane.

The total backplane bandwidth should be high enough in order to accommodate the expected traffic on a switch. Now the total bandwidth of all interfaces is usually higher than the available backplane bandwidth, because a simultaneous usage of full bandwidth on all ports is highly unlikely. A balance is struck providing enough backplane bandwidth to accommodate most situations.

I hope this has been helpful!

Laz

Hello. Why are layer 3 switches used in access layer? (according to the first chart of recomended models).
Thank you!

1 Like

Hi Eilu,
Hope you are doing good:smiley:
(according to the first chart of recomended models).
which chart you are talking? can you paste the screen shot here ? actually i am trying to find out your answer but not able to understand which chart you talking.

Thanks & Regards,
Arindom

1 Like

I believe he is talking about this chart.

switch%20chart

There are layer 3 switches in the chart but this is because those are the “lower end models” that Cisco has to offer. They can run as layer 2 switches and not utilize the layer 3 functionality. As switches have gotten more powerful layer 3 functionality has been easier to put into a switch. A big consideration of what switches you might use in a network is the size of said network. A tiny network might only need 2960-X’s for the core, distribution, and access layer. I hope this helps!

Thanks,
Scott

1 Like

Hello Eliu

You can use Layer 2 switches at the access layer, and that would be fine. However, keep in mind that if you have multiple VLANs running on a single switch, any communication between those two VLANs will have to be sent up to the distribution layer to be routed and then back to the access layer switch once again, to reach the destination.

Now for some networks, that’s fine, however, if there is much inter-subnet communication in a network between devices that are connected to the same switch, it may be worthwhile (although more expensive) to provide that routing within the access switch itself. It all depends on the applications being used and the expected traffic between such devices.

Remember, if your uplinks are being used to transfer data between two devices on separate VLANs on the SAME switch, you may be using up valuable bandwidth of the routing will have to take place at the distribution layer. It all depends on your network design, uplink bandwidths as well as the expected traffic.

I hope this has been helpful!

Laz

1 Like

Hi Rene
If we have multiple vlans environment , then what about SVI creation. SVI will be created on each Distribution layer switch or only single distribution layer switch.

Hello Muhammad

The VLANs that are used by each of the distribution and access switches will be created within them. However, the SVI for each VLAN should exist on only one device, typically the distribution layer switch that serves the VLAN. That way there is only a single location where inter VLAN routing takes place. Now if you use a redundant gateway protocol such as HSRP, then you will have two devices (or more) with an SVI interface for routing purposes, but they will share the responsibility using a virtual IP. You can find out more about HSRP at this lesson:

I hope this has been helpful!

Laz

1 Like

Hi Rene,
Could you give an example of commands sequence for calculation of a switch fabric capacity? e.g. Cisco 2960 series. Thanks in advance.

Hello Vadim

There are no commands that you can use to determine the internal capacity of a network device. The best way to find out what you need is to look at the datasheets of the devices in question. Specifically, the 2960 series can be found here:


Based on this document, you can look at two values that will give you information about the fabric speeds:

  1. Forwarding Bandwidth - This is the maximum bandwidth in bits per second that the the switch can forward simultaneously internally
  2. Forwarding rate - this is the maximum number of packets that can be transmitted in packets per second (or millions of pps). This takes into account the processing power needed to switch each individual packet.

I hope this has been helpful

Laz

Hello Rene/Laz
I have some questions and I am going to use the below diagram as the reference.

In this topology, I have three different sites that are geographically located at three different places. As you see in the diagram, every single site has their own internet connection and they are hosting their public facing servers by using their own Public IP addresses. Now the requirement is to have internet redundancy for every single site. For instance, if the internet connection at
SITE -C goes down, it should be able to use SITE -B or SITE -C as the backup path. In this situation, how can I have redundancy for public facing servers among these three sites. Meaning If the internet connection goes down at
SITE -C, still internet traffic should be able to access the public facing websites that are being hosted by SITE -C by using SITE -B or SITE -C internet connection. Please shed some light on it.

Thank you so much in advance.

Best Regards,
Azm