STP topology change

aueddonline

I have been a bit confused over how long it taks to flush out the bridging table after a TC has been flagged by the root in a configuration BPDU. I thought it was shorted from 300 seconds to 15 seconds but here in the BCMSN cert guide

The TCN flag is received from the Root, and both Catalysts B and C shorten their bridge
table aging times. This causes recently idle entries to be flushed, leaving only the actively
transmitting stations in the table. The aging time stays short for the duration of the Forward
Delay and Max Age timers.

It mentions the MAX age, why would the MAX age time be added?

bighornsheep

I read your post and was a bit confused myself as well, but this slipped my mind until this weekend. This is what I think:

The purpose of a TCN is to inform nodes in the spanning tree topology that change as occured. Either a port is down, or new switch has joined. Remember the objective of the spanning tree protocol is not to create the ideal/best topology but rather a loop-free one. The underlying concern is how frames are forwarded from one switch to another; which relies on the MAC table.

You are right about the reduction to 15seconds (or whatever forward delay is configured as) for the MAC aging timer, this is to speed up the switch's MAC table re-creation so that the new topology can take effect quicker with new forwarding paths being used.

Why the BCMSN guide talkes about Forward delay + MAX age is referring to how long the aging time becomes the forward delay timer of 15seconds; for a period of forward delay + MAX aging timer (default 15+20=35seconds), the MAC aging timer will be the forward delay (default of 15seconds), after the 35seconds, MAC aging goes back to 300seconds.

The 15+20 is derived from how long it takes for the switch which sent the TCN to react to the Topology Change Acknowledgement in the new BPDU from the root bridge.

Source: Kennedy Clark's LAN Switching, Chapter 6 - Understanding Spanning Tree.