thanks for these links

but i need to have some debugging tools for find the culprit switch annoucing as new root

will debug spanning-tree do that ?

which debugs commands or others logs can i set in order to catch it ?

JYP

no wireless , but mixed switches vendors..

with lot of loops

JYP

Opps.. First thing enable " STP rootguard"/" STP loop guard " " UDLD " on all the Cisco switches.

1. Use " Show spanning-tree details ", this command will tell you the Toplogy change notifications on the switch ports coming. If there are TCN it will tell you the last TCN generated by the respective mac-address.

2. Try tracing the mac-address from where the TCN is getting generated by switch and check if STP working properly.

3. Make sure that you are running a compatibe STP between the Cisco and other switches like PVST+ or RPVST+.

4. Debugging the STP events would be helpful to see all the related information.

-amit singh

thanks for these informations

for my understandinf some swithes are running Mstp and others STP..

so even supported seems not the better configuration in production

is it supported by cisco ?

JYP

i would say Mstp and single STP configurations

on same network

is it a supported configuration by cisco ?

JYP

Well it does work and is supported in end-end Cisco environment like you are running MSTP on some switches and CST ( Common spanning-tree IEEE 802.1D) on some of the switches. Its hard to predict how would it behave in a multi-vendor environment although the expected behaviour is that it will run fine.

Once MSTP sees that there are some CST BPDU's being received it marks those ports as MST boundary region ports and runs CST on the top of it to be compatible with the other switches.

http://www.cisco.com/en/US/tech/tk389/tk621/technologies_white_paper09186a0080094cfc.shtml#standard

-amit singh

Usually what i do is manualy set the priority lowest on the one i want to be root slightly higher on what would be the secondary. Then on all of the devices on the network set it one up from the default. For troubleshooting usually do the sh mac-ad | in then follow it through the network to were it resides. Once at that switch usually i do a sh spammingtree root. This will tell you were to go from there. It should show your root server and hopefully all go out the same interface with the same cost.

If you have different roots for different vlans this method will work as well.

Most times in the past this has been cause by a rogue wire looping the network on certian vlans.

Using the above posted commands will help in disableing those ports for you automaticly. you can enable errdiable recovery for these.

on root switch i can so a topology change counters incremented but nor address

does it mean that is only on switch come down/up nor asking as new root ?

explaination welcomed

JYP

VLAN0001 is executing the ieee compatible Spanning Tree protocol

Bridge Identifier has priority 12288, sysid 1, address 0014.1c9c.d500

Configured hello time 2, max age 20, forward delay 15

We are the root of the spanning tree

Topology change flag not set, detected flag not set

Number of topology changes 2064 last change occurred 1w5d ago

from GigabitEthernet4/46

Times: hold 1, topology change 35, notification 2

hello 2, max age 20, forward delay 15

Timers: hello 1, topology change 0, notification 0, aging 300