07-26-2019 02:25 AM
Hi all,
I was wondering about the correct approach to update a vPC pair of 3172TQ switches from a 6.x to a 7.x version.
According to the 7.x software release notes we can go directly from 6.x to 7.x without a intermediate step.
So I copied the .bin file to the bootflash of both switches, checked the file with "sh incompatibility" and "sh install all impact".
With the "sh install all impact" I get an "SRG file error" but accroding to some other posts it seems to be a known issue.
Then I started the installation process on the primary with "install all nxos file_name.bin".
The vPC partner was reachable and accessable all the time and the ports the cluster is connected to were UP/UP all the time too(on the secondary), but the cluster was not reachable anymore. Only after the primary came back up and was reloaded a second time, then the cluster was reachable again.
After I started the update on the primary, it went through the various steps and reloaded. A couple of minutes later, even though the switch was not pingable yet, the vPC formed and I could see the primary from the secondary via CDP.
I took the primary about 40 minutes to become reachable again.
So my questions are:
Any idea why the cluster was not reachable anymore even though the secondary looked alright?
Did I do something wrong? Do I need to use another command?
And another question is, does traffic usually run via the vpc-peer-link? I noticed that the two uplink ports on the primary are blocked and the peer-link ports are in FWD state.
Thanks for any help!
07-26-2019 08:39 AM
Since we do not know the you logs of upgrade, we can not comment what went wrong.
If you configured vPC, all the links in the vPC is active here. But we can only confirm what is wrong by looking your config
also some information output as below
show vpc
show spaning
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide