hi,
I have strange problem - connection between two switches if lost forewer after reboot of the root switch.
Details:
It started with not formed Etherchannel and lost connection between the two switches
or with topology loop followed by broadcast storm.
After some investigations, I was able to reproduce the problem with following basic cfg:
Switch 1
- Cisco Catalyst WS-C2950-24
- SW: IOS (tm) C2950 Software (C2950-I6Q4L2-M), Version 12.1(12c)EA1, RELEASE SOFTWARE (fc1)
- HW: cisco WS-C2950-24 (RC32300) processor (revision H0) with 21002K bytes of memory
Switch 2
- Cisco Catalyst WS-C2950-24
- SW: IOS (tm) C2950 Software (C2950-I6Q4L2-M), Version 12.1(12c)EA1, RELEASE SOFTWARE (fc1)
- HW: cisco WS-C2950-24 (RC32300) processor (revision H0) with 21002K bytes of memory
port fa0/23 of Switch 1 connected by crossover cable to port fa0/23 of Switch 2
port fa0/24 of Switch 1 connected by crossover cable to port fa0/24 of Switch 2
all parameters set to defaults using
#write erase
management interface Vlan1
for debuging purposes:
#service timestamps debug datetime msec
#service timestamps log datetime msec
#debug spanning-tree events
under normal conditions the Switch 1 is becoming spanning tree root, as it has lower MAC address
now the problem:
when the Switch 2 (NOT root) is restarted (power off/on) then everything goes well:
Switch 1 is root, fa0/23 FWD, fa0/24 FWD
Switch 2 sees root on fa0/23, fa0/23 FWD, fa0/24 BLK
but when the Switch 1 (root) is restarted then there is no connection between them (ping) and:
Switch 1 is root, fa0/23 FWD, fa0/24 FWD
Switch 2 is root, fa0/23 FWD, fa0/24 FWD
seems to be a bug in how STP implemantation reacts to interface/link down events
but as I'm not expert, ...
any idea?
thanks
ps
