HSRP - long routing recovery time?

[fcatalao]

Netwotk as follows: 1 6500 & 1 6009. The 6009 has 2 SUP1A with PFC/MSFC2 & the 6509 has 2 SUP1A with PFC/MSFC.

HSRP between them. Where one is active, the other is standby and vice-versa. In router-16 (both switches) all vlans are standby. Access switch is a 3524, with one trunk to 6009 and the other to 6509. The two 6000 are connected by two etherchannel/trunks (1 in sup_slot1, other in sup_slot2). When I switch the 6009 off there is a long delay (4 min) before routing recovers. If I switch the 6509 off routing recovers after a few seconds.

WHY?

Also in my 6509 I keep getting

"%STANDBY-3-DUPADDR: Duplicate address 10.1.3.252 on Vlan21, sourced by 0002.7e22.8c9c"

the documentation mentions that this will not intefer with HSRP working correctly. If disconnect all cables & leave 6509 as stand alone switch this message keeps rolling in.

[lgijssel]

A timeout of 4 minutes points to a mac-table timeout.

From the trunks from 3524 to the 6xxx, only one will be active per vlan. The other one is in blocking state. I expect that the described behaviour will turn over to the other switch if you alter the active uplink to the core.

There is too little info for the dupaddr problem. Does it reappear if the switch is turned off and back on without cables?

[fcatalao]

With respect to the dupaddr problem, yes if I disconnect all cables the problem still exists.

I´m not sure on what you mean with altering the active uplink to the core. What I have done is configure 6009 being the root bridge for half the vlans and secondary root bridge for the other half. On the 6509 I did the same but inversely. All ports are in the forwarding state except for one port in the access layer switch which is in the blocking state.