EtherChannel long fail-back issue between Cisco and other Vendors
I was doing some testing at a customer with their existing equipment (Nortel and Extreme) and a demo Cat 6500 . When I set up link-aggregation (Gig and Fast EtherChannel on the Cisco side) between the switches, connectivity and traffic redistribution were not an issue on initial failover. Where I saw an anomaly was when I reconnected previously used links.
What was happening was that the Nortel and Extreme switches would start to send traffic down a restored link as soon as they saw link establishment on that port. The Cisco 6500 on the other hand, would not send traffic down this port until it added this port to it's bridge group - typically 5-15 seconds depending on how the Channel group was set. The net effect of this is that the other switches would redistribute traffic down a link that was basically a black hole for 5-15 seconds and connectivity was lost for that period.
Just for everyone's benefit, if the Cisco Channel group was set to "Auto or Desirable", the fail-back time was about 15 seconds. If it was set "On" (not listening for PAgP packets), the fail-back time was about 5 seconds. FYI - Spanning-Tree was disabled on these links.
I was looking for a way to reduce the 5 second fail-back time to about 1 second. The customer is interested in the Cat 6500 but wants the failover to happen quicker with the equipment that they already own. Nortel to Extreme fail-over and fail-back happen in less then 1 second. Any real world experiences would be greatly appreciated.
We are pleased to announce availability of Beta software for 16.6.3. 16.6.3 will be the second rebuild on the 16.6 release train targeted towards Catalyst 9500/9400/9300/3850/3650 switching platforms. We are looking for early feedback from custome...