What happes when cst root fails?

[Packet-Ghost]

Hi,

We have several sites connected with L2 links and we run variuos flavours of STP on them.  We are now in the process of reconfiguring two sites to prepare them for new L2-connections to a third and forth connection that will hold the CST Root.

Basically we are running several MSTP regions and we need to control the CST (And IST's) by using the stp priorty setting.

In the two sites in question, they have enabled MSTP, but not really configured it properly, so everything is running in Instance 0 (IST). We've managed to get the region-name and rev. correctly on all of them, so they now all agree on the CST and IST. So far so good...

We have the root in site A with a pri of 10 (40960) and then we have a switch with pri 11 (45056) in site B. All other switches are running at pri 12 or higher (we've put in place root-guards and bpdu-blocking) to make sure..

The quesion is: If we have an outage on site A, meaning the CST Root fails, what will happen? I know off course that my pri 10 switch will be the new root, but will this cause a new STP Recalculation meaning a "long" outage? Or will it simply be a "failover-situation" meaning we will only loose a few pings (the same as if a link goes down) ?

What is the consequence when CST root fails?

Thanks!

K.

[paulgear]

Hi K,

If the root fails there will always be a longer outage than if a non-root path fails. However, by tuning your timers you should be able to reduce this to a minimum.

I would suggest setting your root switch's priority to something lower than 8, because if someone brings along a new switch that is unconfigured, it will actually take over from your root, because the default setting is 8. I usually set the priority on my preferred root to be 0, that way i know nothing else is going to come along and out-vote it.