Hi,

Can you describe your problems in more details?

Usually there are no issues between R-PVST and other versions of STP, except probably MST, that needs some more configuration

Here's the set up:

Access layer switch A1 (Dell 6224) is dual homed to 2 Cisco Distribution switches, D1 and D2. And D1 and D2 are connected to each other. Typical triangle topology with all vlans trunked...

 

When testing, I was PINGing D1 from A1 and then disconnecting the active link (A1-D1), letting it failover to the backup (A1-D2). Failover was fast, I would only drop ONE PING. Then I reconnected what was the active link (A1-D1). When I did the reconnect, I had to wait for the Cisco switchport on D1 to go through all the STP states before forwarding traffic, so I would drop PINGs for about 40 seconds. (Thats the problem, very slow reconvergence when the failed link comes back online)

 

This wouldnt happen if the access layer was a Cisco switch. The failback upon reconnect would have been immediate because Uplinkfast and BackboneFast config BPDUs would be exchanged, informing the D1 switch that its part of an Uplinkfast group. But since the access switch is a NON-Cisco switch, they dont "talk" to each other about uplinkfast. However, they should be talking to each other with 802.1w and negotiate the proposal and agreement. Something is wrong.

 

You have brand new switches but maybe there's newer software version, try to update if possible. For your indormation: Dell PC software is very raw and sometimes works awful.

When you firstly connecting Dell to Cisco does link come up quickly?

Did you get information about STP convergence through debugging?

For some reasons Cisco (or Dell) can't talk with R-STP and fall back to STP. Can you check link type on inter-switch interfaces while reconverging? Is it p2p or something like p2p peer (stp)/p2p bound (stp)?

 

Have a look at this doc - Dell-Cisco STP Interoperability and Recommendations v2.0.pdf at:

supportforums.cisco.com/.../DOC-16177