Nexus 93180-FX3 VPC HSRP traceroute 2 hops similar

[Jeanvanlord]

Hello,

We have receved 2 configured new Nexus 93180YC-fx3.

CORE01: Link OSPF on ETH1/1 (Connected and OK)

CORE01 and CORE02 VPC activate.(L3 option activate)

VLAN 253 with HSRP configured.

When a Linux serveur traceroute and connected to Core01:

1 10.24.253.2 (10.24.253.2) 0.389 ms 0.367 ms 0.411 ms
2 10.24.254.4 (10.24.254.4) 0.392 ms 0.337 ms 0.314 ms
3 10.10.32.145 (10.10.32.145) 2.674 ms 2.739 ms 2.663 ms

When connected on the core02:

1 10.24.253.2 (10.24.253.2) 0.849 ms 0.826 ms 0.826 ms
2 10.24.253.2 (10.24.253.2) 0.362 ms 0.406 ms 0.447 ms
3 10.24.254.4 (10.24.254.4) 0.407 ms 0.344 ms 0.286 ms
4 10.10.32.145 (10.10.32.145) 2.626 ms 3.283 ms 3.294 ms

What can explain this double identical Hops ?

Bad configuration or normal due to HSRP and VPC ?

-------

Primary HSRP:

interface Vlan253
no shutdown
no ip redirects
ip address 10.24.253.2/24
no ipv6 redirects
ip router ospf 2 area 0.0.0.0
hsrp 53
ip 10.24.253.1

--------

[MHM Cisco World]

Can you more elaborate about your issue

[Jeanvanlord]

[Jeanvanlord]

Thanks for  your answer.

As show in the picture.

When server Connected to Core02

1 10.24.253.2 (10.24.253.2) 0.849 ms 0.826 ms 0.826 ms
2 10.24.253.2 (10.24.253.2) 0.362 ms 0.406 ms 0.447 ms
3 10.24.254.4 (10.24.254.4) 0.407 ms 0.344 ms 0.286 ms
4 10.10.32.145 (10.10.32.145) 2.626 ms 3.283 ms 3.294 ms

We can see 2 time: 10.24.253.2

And when i make test, The server say that identify Asym Trafic.

But not when connected on the first.

1 10.24.253.2 (10.24.253.2) 0.389 ms 0.367 ms 0.411 ms
2 10.24.254.4 (10.24.254.4) 0.392 ms 0.337 ms 0.314 ms
3 10.10.32.145 (10.10.32.145) 2.674 ms 2.739 ms 2.663 ms

I know that HSRP work in actif/actif on VPC but that provocate this Asym error ?

Thanks

[MHM Cisco World]

This is traceroute 

Server to nsk-1 ( hsrp vip is use by both nsk since it active/active)

1 10.24.253.2 (10.24.253.2) 0.849 ms 0.826 ms 0.826 ms

Here 

Server to nsk1 to nsk2

Again since it vpc hsrp active/active 
2 10.24.253.2 (10.24.253.2) 0.362 ms 0.406 ms 0.447 ms

Here 

Server to nsk1 to nsk2 to core SW
3 10.24.254.4 (10.24.254.4) 0.407 ms 0.344 ms 0.286 ms

Here the end point 
4 10.10.32.145 (10.10.32.145) 2.626 ms 3.283 ms 3.294 ms

So this normal behavior of vpc nsk I think.

I will check how we can remove the nsk2 from path.

MHM

[MHM Cisco World]

Since it Single leg trt use vlan SVI of nsk1 as gw of server and check.

Note:- nsk1 is nsk server direct connect to it

MHM

[Jeanvanlord]

Hello,

Thanks for your answer.

Test Make:

Remove interface vlan config from Nexus 2 but server always connected on it:

1 10.24.253.2 (10.24.253.2) 0.389 ms 0.367 ms 0.411 ms
2 10.24.254.4 (10.24.254.4) 0.392 ms 0.337 ms 0.314 ms
3 10.10.32.145 (10.10.32.145) 2.674 ms 2.739 ms 2.663 ms

Nexus 1 as Gateway :

1 10.24.253.2 (10.24.253.2) 0.289 ms 0.267 ms 0.311 ms
2 10.24.254.4 (10.24.254.4) 0.292 ms 0.335 ms 0.311 ms
3 10.10.32.145 (10.10.32.145) 2.401 ms 2.712ms 2.200 ms

Roll back: Interface Vlan back + HSRP + server on Nexus2

1 10.24.253.2 (10.24.253.2) 0.849 ms 0.826 ms 0.826 ms
2 10.24.253.2 (10.24.253.2) 0.362 ms 0.406 ms 0.447 ms
3 10.24.254.4 (10.24.254.4) 0.407 ms 0.344 ms 0.286 ms
4 10.10.32.145 (10.10.32.145) 2.626 ms 3.283 ms 3.294 ms

Asymm detection:

[MHM Cisco World]

Server is connect via one leg to nsk vpc pair.

So if we use hsrp vip both nsk can forward traffic destiantion to this IP.

But this IP also use for reaching external subnet connect to nsk-2 that why the IP appear twice.

What I suggest is making server (since it single leg connect) use vlan svi of nsk-1 as gw 

This make tracroute not show same IP twice.

MHM