Simulating UniDirectional Link Failure

[d.hodgson]

Hi folks,

I may be losing the plot but can I simulate unidrectional link failures using fibre to help my CCNP studies? I thought it would be a simple matter of pulling out just the rx on one of my switches but that brings the int down. Is this how it's supposed to work?

I want to do this to look further into udld.

thanks

Dave

[Nagendra Kumar Nainar]

You can use PACL to simulate unidirectional link failure. UDLD uses 0100.0ccc.cccc as destination MAC address.

So configure a PACL denying this mac address and apply the same to the interface on which you need to simulate the link failure.

HTH,

Nagendra

[d.hodgson]

Thanks Nagendra,

I'll look up PACL.

Now what about the fibre failure I simulated, is expecting that to be a unidirectional link failure a wrong assumption?

ta

Dave

[thibaultm]

That is a brilliant idea, I wanted to reproduce a unidirectional failure of a fiber in a lab CML or GNS3.
Until I stumbled upon this page from a proficient Cisco employee :
https://cisco.goffinet.org/ccna/securite-lan/cisco-port-acl-pacls-vlan-acl-vacls/

I translate in English the main points :
- PACL only filter incoming traffic
- PACLs CANNOT filter L2 protocol frames like : VTP, CDP, DTP, PAgP, UDLD and STP
- PACLs are supported only on Hardware

With all that it doesn't seem possible to simiulate in a simulated Lab
if so it is simpler to build a HW network and then unplug 1 direction only fiber
in front of a RSTP blobked port by RSTP.

[h.parsons]

It is my understanding that your assumption is right. One thing to note is that both switches will have to establish a bidirectional state once before UDLD will shutdown an interface. Try connecting them and do a "show udld neighbors"

[shaktik]

I know its quite late for the answer, but just for other people visiting here.. I was just thinking that attaching the Rx to a higher voltage source should work out since it should complete the loop