11-22-2011 07:45 AM - edited 03-07-2019 03:32 AM
Hello Everyone:
As shown in the attached diagram. When I ping the Virtual IP on the server I lose a ping over every 6 in the first 2 configurations.
When I replace the Nexus 5K by a Catalyst 3560, the result is smooth.
As you can see in the second topology I am connecting both 4K to the same N5K in order to get the VPC peer link in the equation but still no success.
Any ideas appreciated.
Note: 4K does not support VPC natively.
Something I observed in the first two topologies is that the MAC address for the server as seen on the NEXUS flaps constantly at a rate on 1 evry 7 sec(aprox.)
For example in topology 2:
When I do a show MAC I will see MACX being learned from port 16 and it will bounce to port 15 for about 1 sec ang go back to port 16.
Thanks for any input.
11-22-2011 08:20 AM
Like you said, the N4K doesn't have vPC support and your first and second diagrams are not supported.
The question is do you want additional bandwidth with active-active? Or is there any other failure scenario that your server wouldn't failover correctly?
Regards,
jerry
11-22-2011 08:48 AM
Thanks for your response:
LACP was not configured on the server(IBM BLADE).
NIC teaming was configured with one active NIC and one standy NIC and still the ping was timing out every 6 request as shown below in the second topolgy.
The primary goal is to have NIC teaming either in standy mode or load-sharing mode on the(outound) working.
I am suprised that it does not work on topology 2 as there is only one switch as parent.
Pinging 10.130.100.149 with 32 bytes of data:
Request timed out.
Reply from 10.130.100.149: bytes=32 time=43ms TTL=122
Reply from 10.130.100.149: bytes=32 time=21ms TTL=122
Reply from 10.130.100.149: bytes=32 time=22ms TTL=122
Reply from 10.130.100.149: bytes=32 time=27ms TTL=122
Request timed out.
Reply from 10.130.100.149: bytes=32 time=19ms TTL=122
Reply from 10.130.100.149: bytes=32 time=22ms TTL=122
Reply from 10.130.100.149: bytes=32 time=20ms TTL=122
Reply from 10.130.100.149: bytes=32 time=22ms TTL=122
Reply from 10.130.100.149: bytes=32 time=25ms TTL=122
Reply from 10.130.100.149: bytes=32 time=23ms TTL=122
Request timed out.
Reply from 10.130.100.149: bytes=32 time=34ms TTL=122
Reply from 10.130.100.149: bytes=32 time=22ms TTL=122
Reply from 10.130.100.149: bytes=32 time=22ms TTL=122
Reply from 10.130.100.149: bytes=32 time=23ms TTL=122
Reply from 10.130.100.149: bytes=32 time=130ms TTL=122
Request timed out.
Reply from 10.130.100.149: bytes=32 time=36ms TTL=122
Reply from 10.130.100.149: bytes=32 time=41ms TTL=122
Reply from 10.130.100.149: bytes=32 time=27ms TTL=122
Here is where the MAC is flapping:
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/15
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/15
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/16
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/15
NX5596-1.CTB# sh mac address-table | inc 0010.1870.7710
* 1312 0010.1870.7710 dynamic 0 F F Eth1/15
Thanks
11-22-2011 09:59 AM
I am not a server guy so I can't really comment on NIC teaming (active-standby) configuration. From my previous experience with an customer running N4K, active-standby NIC teaming should be okay.
On the other hand, you should not configure load-balancing on the server side because it will cause MAC for sure.
Regards,
jerry
11-22-2011 10:08 AM
Thanks Jeye:
That is what I though too. That is standby active should have no issue.
If I get the Standy-ACtive working I will be more that happy. We have now a Case open with Cisco and also on the server side.
Thanks
11-22-2011 10:19 AM
Not a problem.
Regards,
jerry
03-05-2014 01:32 AM
Hi Shummoogum,
I actually have the same issue. How did you resolve it?
Thank you very much for your help.
Oscar
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: