Layer 2 loop prevention in uplink switches where catalyst switches is being used and in downlink nexus switches are being used with vPC to server

[olly ahmed]

Suppose I have two nexus switches in downlink which is configured as vPC to connect a server and server is also using nic teaming. Both nexus are also connected to both uplink switches for redundancy. My question is as far as I know for vPC both nexus will forward traffic of the server to uplink, but how uplink catalyst switches prevent loop ? Would appreciate if anyone can describe this fact for vPC working properly as well as peer-link and keep-alive link fails between two nexus.

[olly ahmed]

Hi ! 

Is there anyone to help me to understand the above issue ?

[devils_advocate]

Hi

How are the nexus' configured in terms of the upstream switches?

Are the upstream also Nexus?

Thanks

[olly ahmed]

Hi devils,

As i mentioned earlier that upstream switches are catalyst not Nexus.

[ahmedshoaib]

Hi;

It's total depend on your design and network requirements.

If the upstream link connected to Catalyst Switch is L2 you can configure vPC (which will provide you double of bandwidth & redundancy).

If the upstream link connected to Catalyst switch is L3 you can configure point-to-point and use peer gateway feature on nexus which will also provide redundancy and utilized both links for traffic forwarding.

Thanks & Best regards; 

[olly ahmed]

Hi Shoaib,

Thanks for your reply. Please be informed that upstream link connected to Catalyst Switch is L2. I just wants to know how catalyst will prevent loops ?

[ahmedshoaib]

Hi;

If you configure Nexus upstream link as a vPC & Catalyst Switch L2 port-channel then there will no loop.

Thanks & Best regards;

[olly ahmed]

Hi Shoaib,

Thanks for your reply. Would appreciate if you please explain the loop prevention mechanism for the attached diagram. Where two layer 2 Catalyst switches are at uplink.

[devils_advocate]

An example of this is here:

https://darow.wordpress.com/2011/08/31/cisco-nexus-5596s-with-redundant-uplinks-to-catalyst-6509-cores-using-vpc/

In order for STP to take care of the redundant link however, you would need a connection between your Catalyst switches in this diagram. 

As these are L2 links, you would need to be careful of your root bridge placement.

[olly ahmed]

Yes I have connection between two catalyst switches but forget to add I the diagram. 

[ahmedshoaib]

Hi;

You have 2 option:

1 - In Case of VSS mode you can create a single port-channel on Catalyst switch & vPC on Nexus side (all 4 ports be in single vPC). Which will provide you 4 times of bandwidth b/w nexus & catalyst switch with redundancy.

2. In case of normal you can create 2 port-channel on catalyst switch side (1 port-channel for each switch) while on nexus you also need to configure 2 vPC (each vPC for links going for catalyst switch). Which will provide you 2 times of bandwidth b/w nexus & catalyst switch with redundancy.

You need to configure STP in such a way Catalyst SW-1 as root primary & Catalyst SW-2 as root secondary. Which will provide you loop avoidance mechanism in your network.

Thanks & Best regards;

[olly ahmed]

Thanks for your reply. Can you share a study link so that I can study deeper regarding this issue.

[ahmedshoaib]

Hi;

Please find the below link "Design and Configuration Guide: Best Practices for Virtual Port Channels (vPC) on Cisco Nexus 7000 Series Switches".

http://www.cisco.com/c/dam/en/us/td/docs/switches/datacenter/sw/design/vpc_design/vpc_best_practices_design_guide.pdf

Thanks & Best regards;