Here's another one I encountered in our lab ... maybe it's something I misunderstood, looking for some enlightenment ...
OK, core setup consists of two 4507 SupV-10GE, connected through the TenG interfaces in a REP ring. Works fine so far. Connected to those two switches (let's say c1 & c2) are four ME3400 switches, all of which going through their GigE ports as a 2G etherchannel (with LACP). Ports are configured as trunks. The other ports on the ME3400 will be used for a MAN application, interconnecting customer sites via QinQ. Now, network functions are fine so far. I thought it might be an added value to the customers and added functionality if the 4500's would tunnel stuff like the STP packets, so I enabled l2protocol tunneling for STP, VTP and CDP (the latter mostly for proof of concept). This seemed to work fine, as now the CDP adjacencies for the ME's were the other ME's and not the 4500 anymore (though every neighbor appeared 4 times, two port for both port each). But then I noticed that one of the switches had both its uplink port disabled. shut/no shut resulted in them being re-disabled due to "misconfigured channels". As I couldn't find any difference (or fault) in the switch's configuration, I rebooted it, which resulted in both uplinks being active again - only to see the uplink ports on yet another switch going down, with the same error message. Rebooting THAT switch yet again led to a stable network, with all ME switches being up and running fine. cdp neighbors were fine, communication between all switches were also fine. Until I decided to reboot another of the ME switches - when it came back up, the uplink ports of the second switch went down again. So, what am I missing here? Tnx, garry _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
