Spanning tree was set to off by the installation engineer. But I am having odd problems on my switch stack (a mix of HP and Aruba) that hit as spanning-tree storms and BPDU starvation messages.
I only have one interface on my 7210 contoller connected to my LAN. My controller has 2 VLANS that are not tunneled. These are 1, where my controller's management interface is, and 114, which is where the controller hands out IP's to the AP's via DHCP (also handled by the controller). On each AP port, I untag it to VLAN 114 on the switch (HP) or set the native VLAN to 114 (Aruba access port). The AP's (I learned the hard way) can't have uplink VLAN set, so those are all zeroes.
The Aruba wifi works fine, it's just that:
a) I get daily or twice daily switch spanning tree problems, and
b) If I disconnect the controller interface for a few seconds, it reliably clears my switch lockups.
My HP switches are doing MSTP and I have VLANS 1,110,111,112,113,114,115. Something is convincing them to go through CST election processes. I have set every non-uplink port to BPDU Protect and portfast/admin-edge-port, yet these problems persist.
#7210