3

There are two trunk links with the same configuration connecting two switches. Currently they are independent links and one is blocked with spanning-tree. I would like to convert these to a single port channel without any down time. Is this possible? What is the best procedure for doing so?

These are Cisco, but I think the concept would apply to any switches that support the same function.

What I've come up with so far:

  1. Shutdown the STP blocked interface, on the core side
  2. Add both interfaces on the down connection to the channel group
  3. No shutdown the interface that was shut down in step 1 (STP will block again)
  4. Shutdown the interface with STP forwarding active (STP will start forwarding on the port channel)
  5. Add the interfaces on the down connection to the channel group
  6. No shutdown the interface that was shutdown in step 4

The problem is that step 4 here creates about 1 second of outage. Is there a way to do this without any outage?

  • 1
    You will lose some frames, no matter what you do (STP will need to converge), and if you do it incorrectly, you can end up with err-disabled interfaces that you must shut down then restart, anyway.. The safest thing to do is shut down both physical interfaces on both switches, add the physical interfaces to the channel group on each switch, and then bring up the physical interfaces on both switches. – Ron Maupin May 30 '18 at 21:57
  • 1
    Which version of spanning-tree are you using (or better yet provide the output of show spanning-tree summary)? – YLearn May 31 '18 at 4:06
  • rapid-pvst Switch is in rapid-pvst mode Root bridge for: none Extended system ID is enabled Portfast Default is disabled PortFast BPDU Guard Default is enabled Portfast BPDU Filter Default is disabled Loopguard Default is disabled EtherChannel misconfig guard is enabled UplinkFast is disabled Stack port is StackPort2 BackboneFast is disabled Configured Pathcost method used is short – Yanzzee Jun 1 '18 at 19:50
2

Short answer: no.

While the STP handover in step 4 or step 3 is usually faster than 1 second there's always a gap. Make sure you're using RSTP, MSTP or RPVST though. The original and largely obsolete STP is slow on convergence.

One tiny thing might reduce the gap a little further: before step 3, decrease the trunk's STP priority value below that of the yet active port. This will already initiate the handover in step 3, so it doesn't need to be a shutdown reaction but a more 'flowing' process.

As this seems to be very delicate connection I'd stage the exact process in a lab before doing it in production.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.