cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1764
Views
5
Helpful
6
Replies

VPC Loop Prevention - BUM Traffic from Orphan

OBD
Level 1
Level 1

Hello everyone ,

 My question is a bit different about VPC loop prevention. We all know that , traffic from remote vpc member is can pass towards to vpc peer via peer-link but it is not allowed to exit from vpc member ports because of loop prevention.

 But  if BUM traffic comes from orphan port to local vpc  , what will be VPC peer's behavior ? Because normally , traffic from remote orphan is allowed to pass through peer link and can exit via local vpc member.  If it happens , it means vpc member ports would receive duplicate frames.

 So do you have any idea about it ? Or am I wrong about  "traffic from remote orphan is allowed to pass through peer link and can exit via local vpc member" part ?

Thank you

 

 

2 Accepted Solutions

Accepted Solutions

Not sure what is the concern here, i do not see any issue here as per your diagram, this is normal and deployment

 

most use case explained in better way here :

 

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

 

https://netcraftsmen.com/how-vpc-works/

 

BB

***** Rate All Helpful Responses *****

How to Ask The Cisco Community for Help

View solution in original post

Yes I know what are you talking about, if BUM forward via local vpc member and forward via other peer vpc member “pass through vpc link” then server have two copy and this is loop.

so in order to prevent that and to emulate two nexus to word as one virtual sw the vpc will never forward traffic receive from vpc link except the case peer vpc member is down.

View solution in original post

6 Replies 6

15356893-vPC.png
Yes from orphan to other orphan connect to other VPC peer the traffic can pass.

OBD
Level 1
Level 1

Hi,

 Actually my question is not about orphan to orphan communication. I wonder behavior of local orphan to remote vpc member communication? (assume traffic is broadcast) Because normally , traffic from remote orphan is allowed to pass through peer link and can exit via local vpc member.So if it happens , it means vpc attached server may get duplicate frames.

 I can't find document about this scenario. But I think vpc peer switch can't let frame to exit via local vpc port when it comes from remote orphan port.

 Since I can't find any document about this scenario ,  I need to be clarified.

 

VPC.PNG



friend, 
from orphan connect to Peer-A of VPC the broadcast will forward to Server connect to Peer-A of VPC through the direct connect link, it not necessary to go through VPC-link and then to Peer-B of VPC and end to server. 
if direct Link between Server and Peer-A down then traffic can pass through VPC-link ->Peer-B->Server 
because Peer-B know that member link is down through control plane between the two Peer.
https://netcraftsmen.com/how-vpc-works/

 

I think you are talking about unicast traffic. My question is about broadcasts traffics. 

 I guess  " Frames received in the vPC Peer Link cannot flood out a vCP Member Port while the remote vPC Peer has active vPC members in the same vPC "

http://www.cciemcgee.com/blog/nx-osvpcs

 So I think I've misinformation that I mentioned in my first question ("traffic from remote orphan is allowed to pass through peer link and can exit via local vpc member"). I don't want to misinform people who read this topics. So you can ignore my "traffic from remote orphan is allowed to pass through peer link and can exit via local vpc member" sentence . It seems it is wrong.

Yes I know what are you talking about, if BUM forward via local vpc member and forward via other peer vpc member “pass through vpc link” then server have two copy and this is loop.

so in order to prevent that and to emulate two nexus to word as one virtual sw the vpc will never forward traffic receive from vpc link except the case peer vpc member is down.

Not sure what is the concern here, i do not see any issue here as per your diagram, this is normal and deployment

 

most use case explained in better way here :

 

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

 

https://netcraftsmen.com/how-vpc-works/

 

BB

***** Rate All Helpful Responses *****

How to Ask The Cisco Community for Help

Review Cisco Networking for a $25 gift card