[f-nsp] Rapid Spanning Tree too slow (like ordinary -not rapid - stp)

Jethro R Binks jethro.binks at strath.ac.uk
Tue Jun 24 15:17:29 EDT 2008


On Tue, 24 Jun 2008, Gunther Stammwitz wrote:

> My config is looking like this:
> vlan 503 name FRA3-FCS by port
>  tagged ethe 49 to 50
>  untagged ethe 4
>  spanning-tree rstp
>  spanning-tree rstp priority 0
> 
> I've already tried to use "spanning tree 802-1w" instead of rstp which
> didn't help.
> 
> this behavior can be seen on bigirons and netirons using jetcore/M4 and 
> irconcore/M3 as well as on the FI4802-switch on the router and the 
> switch-codetrees.

One thing I can tell you is that 802-1w is what you need - that is 'real' 
RSTP.  Confusingly, the 'rstp' in the Foundry is one of the earlier drafts 
for RSTP.  See the documentation.

You should also specify any non-shared media link (ie, not to a hub) as 
point-to-point, for example:

  spanning-tree 802-1w ethe 1/1 admin-pt2pt-mac

And that's about as far as my knowledge goes :)  But it works in the 
situations I've deployed it (BigIron, Super-X) like this, with topology 
change recovery times of 2 to 4 seconds, as expected.

I'd be glad to hear any other optimisations people can suggest.

Jethro.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Jethro R Binks
Computing Officer, IT Services
University Of Strathclyde, Glasgow, UK



More information about the foundry-nsp mailing list