Hello community,
We have seen something bizarre on a Nexus switch. We were having disconnection issues on a 10GE FEX B22 interface that was configured as trunk. On the client side, there is an HP UNIX blade system connected. It has two interfaces that are configured as active/standby trunk interface (an APA, Auto Port Aggregation interface)
On the Nexus side, the interfaces were configured as follows:
interface <1>
switchport mode trunk
switchport trunk native vlan 999
switchport trunk allowed vlan 371,377
interface <2>
switchport mode trunk
switchport access vlan 371
switchport trunk native vlan 999
switchport trunk allowed vlan 371,377
Interface <2> had a logical status of "down" on the HP side, but the interface was physical UP on the Nexus side. The fact that interface <2> sometimes goes to down on the unix side is the main problem.
Now, i have always thought that the command "switchport access vlan 371" has no effect and doesn't really do anything when the interface is configured in trunk like it is. But, when we just removed this line:
interface <2>
no switchport access vlan 371
exit
the interface on the HP UNIX side changed to "UP". Is it possible that this command triggered some sort of soft-reset of the interface, because in my opinion, it has not changed anything, the vlans allowed are still the same, the native vlan is also still the same. For example, might it renegotiate duplex or speed settings (fast link pulses) or anything like that ?
Nexus switch was Nexus 56128P, running 7.2(1)N1(1)
regards,
Geert
