I find myself struggling with network setup. I've read through various forums and I haven't found anything that fits what I'm seeing, which I think should work but is not.
I have 2x dual-port 10GB Intel cards connected to two Cisco 4500X switches running rapid-pvst. Port 0 of each 10GB card (ql0 & ql2) go to switch 4500X-A and Port 1 of each 10GB card (ql1 & ql3) go to 4500X-B.
On each switch I have the same configuration: Te1/5 & Te1/5 associated with port-channel 25, trunk mode, allowed VLANs 150 & 250. On the server side I have 2 LAGGs setup, lagg1 (ql0,ql2) & lagg2, (ql1,ql3) both using LACP. I then created VLANs for the first network, vlan10150 with a parent interface of lagg1 and vlan20150 with a parent interface of lagg2. I then created bridge150 with bridge members vlan10150 and vlan20150 and gave bridge150 an appropriate IP address. I have not yet setup the VLAN 250 interfaces or bridge. Also, switch "A" (connected to lagg1) is configured as the primary root bridge for VLANs 150 & 250 while switch "B" (connected to lagg2) is configured as the secondary.
When I enable the "A" side network ports the port-channels come up and I can do a "tcpdump -i bridge150" on the server and see STP traffic (among other things). I can ping and traceroute to the router and other servers, all traffic passes through the network perfectly, which tells me the lagg is up and the traffic is being appropriately tagged. When I look at ifconfig bridge150 I don't see a STP root id which is a problem even with only one side up. When I enable the "B" side of the network everything goes down hill, a layer 2 loop is created and the network goes crazy. STP is simply not functioning although I can clearly see 802.1W packets coming in from the switch with a tcpdump. To resolve the loop I have the "B" side network ports shutdown.
I tried issuing the following on the command line:
to ensure that STP was running but I get the message:
I'm not sure what I'm missing. Any ideas?
J
I have 2x dual-port 10GB Intel cards connected to two Cisco 4500X switches running rapid-pvst. Port 0 of each 10GB card (ql0 & ql2) go to switch 4500X-A and Port 1 of each 10GB card (ql1 & ql3) go to 4500X-B.
On each switch I have the same configuration: Te1/5 & Te1/5 associated with port-channel 25, trunk mode, allowed VLANs 150 & 250. On the server side I have 2 LAGGs setup, lagg1 (ql0,ql2) & lagg2, (ql1,ql3) both using LACP. I then created VLANs for the first network, vlan10150 with a parent interface of lagg1 and vlan20150 with a parent interface of lagg2. I then created bridge150 with bridge members vlan10150 and vlan20150 and gave bridge150 an appropriate IP address. I have not yet setup the VLAN 250 interfaces or bridge. Also, switch "A" (connected to lagg1) is configured as the primary root bridge for VLANs 150 & 250 while switch "B" (connected to lagg2) is configured as the secondary.
When I enable the "A" side network ports the port-channels come up and I can do a "tcpdump -i bridge150" on the server and see STP traffic (among other things). I can ping and traceroute to the router and other servers, all traffic passes through the network perfectly, which tells me the lagg is up and the traffic is being appropriately tagged. When I look at ifconfig bridge150 I don't see a STP root id which is a problem even with only one side up. When I enable the "B" side of the network everything goes down hill, a layer 2 loop is created and the network goes crazy. STP is simply not functioning although I can clearly see 802.1W packets coming in from the switch with a tcpdump. To resolve the loop I have the "B" side network ports shutdown.
I tried issuing the following on the command line:
ifconfig bridge150 stp vlan10150 stp vlan20150to ensure that STP was running but I get the message:
ifconfig: unable to set bridge flags: Invalid argumentI'm not sure what I'm missing. Any ideas?
J
