Solved: Spanning Tree network unstable

alfonso.cornejo · ‎01-26-2010

Hi,

I'm using rapid-pvstp in my fiber ring and here is problem: i have to main sites, A and B, and a chain of 30

switches from site A to site B, for the main site A i have the vlan 11 and for the main site B the vlan 12, the

switch closer to site A is configured to be the root of vlan 11 and the switch closer to site B is configured

to be the root of vlan 12, the problem that i have is that the network is very unstable, the root bridge is

changing from the right one to other one very often (sometimes every second) and that is causing me that some

ports sometimes are in blocked state and a second later in forwarding state and then blocked again.

It looks that the switches that are far away from the root of each vlan are not

receiving the bpdu or they receive them with bad values so they start to make them self roots or choosing

another root, doing some researching i found this: "When a BPDU propagates from the root bridge toward the

leaves of the tree, the age field increments each time the BPDU goes though a bridge. Eventually, the bridge

discards the BPDU when the age field goes beyond maximum age. If the root is too far away from some bridges of

the network, this issue can occur. This issue affects convergence of the spanning tree."

http://www.cisco.biz/en/US/tech/tk389/tk621/technologies_tech_note09186a00800951ac.shtml#awkward

What

I was thinking is to modify the hello, forward and max age timers on the root switches (root for vlan 11

and root for vlan 12) but i'm not sure what value should i use in each timer.

Any other ideas??

Here i attach the diagram for in order to have a better idea of the scenario.

Thanks in advance for your help!

Giuseppe Larosa · ‎01-26-2010

Hello Alfonso,

your calculations are good.

Anyway I would simply use

hello = 1 sec

max-age= 40 sec

This should give margin also for the chain scenario where one link is broken

Hope to help

Giuseppe

View solution in original post

Giuseppe Larosa · ‎01-26-2010

Hello Alfonso,

you can consider the max-age time as a sort of TTL for the STP BPDU

max-age / hello frequency gives you how far BPDUs are considered valid

so you may need to increase max-age and to reduce hello interval to achieve an higher ratio so that BPDUs can travel correctly

You need to consider what can happen when a failure changes the ring in a long chain (worst case)

This has been discussed in some previous threads

Hope to help

Giuseppe

alfonso.cornejo · ‎01-26-2010

Hi Giusepe,

Right know i have the default Hello 2 sec, Max age 20 and Forward 15, do you think this values can do the job:

Hello 2 sec

Max age 40

Forward 15

???? or is there a formula to calculate them???

I mean because of the fact that i have a chain of like 30 switches

Thanks in advance for your help!

alfonso.cornejo · ‎01-26-2010

I believe this are the formulas:

End-to-end BPDU propagation delay—This value is the amount of time that is necessary for a BPDU to travel from one end of the network to the other end. Assume a diameter of seven hops, three BPDUs that can be lost, and a hello time of 2 sec. In this case, the formula is:

End-to-end_BPDU_propa_delay
= ((lost_msg + 1) x hello) + ((BPDU_Delay x (dia – 1))
= ((3 + 1) x hello) + ((1 x (dia – 1))
= 4 x hello + dia – 1
= 4 x 2 + 6
= 14 sec•Message age overestimate—The purpose of this parameter is to account for the age of the BPDU since origination. Assume that each bridge increases the BPDU message age by 1 sec. The formula is:

Message_age_overestimate
= (dia – 1) x overestimate_per_bridge
= dia – 1
= 6

max_age
= End-to-end_BPDU_propa_delay + Message_age_overestimate
= 14 + 6
= 20 sec

I'm i correct?

This values only have to be changed on the root switches right? for this case on my actual root sw for vlan 11 and my actual root for vlan 12?

Thanks!

Giuseppe Larosa · ‎01-26-2010

Hello Alfonso,

your calculations are good.

Anyway I would simply use

hello = 1 sec

max-age= 40 sec

This should give margin also for the chain scenario where one link is broken

Hope to help

Giuseppe

alfonso.cornejo · ‎01-26-2010

Thanks alot Giussepe!

This values only have to be changed on the root switches right? for this case on my actual root sw for vlan 11 and my actual root for vlan 12?

Thanks!

Giuseppe Larosa · ‎01-26-2010

Hello Alfonso,

yes, sorry for not having noted this

of course configure also on secondary root bridge to have consistency in case of failure of primary root bridge

Hope to help

Giuseppe

alfonso.cornejo · ‎01-26-2010

No problem,

I'll perform this configurations tonight and I'll post how it goes.

Thanks again!

alfonso.cornejo · ‎01-27-2010

Hi Giuseppe,

I did the changes last night and everything is working as it should be!.

Thanks for your help...