ether VIF 2/8 down, reason: Bound Physical Interface Down

[HamR]

Can anyone please advise what this message means and why a host reset causes it?

If I disassociate and re-associate the SP the error goes away, although this isn't much fun when the server boots from iSCSI.

Currently running 2.0(2Q).

[abbharga]

Hi Hamish,

This message indicates the upstream uplink port to be down thus bringing the pinned veth down as well.

Since this is for iSCSI, do you by any change have a l2-disjoint uplinks configured? Seprate uplink for iSCSI network and seprate for regular traffic? If yes, are you allowing the vlans from different uplinks on the same veth, since that can also lead to this situation.

Abhinav

[HamR]

Hi Abhinav,

I suspect the error may be symptomatic of another problem.

All upstream uplinks are appearing up. I've no disjoint-L2 uplinks configured. The frustration is that the server works fine until it is reset. Then I can disassociate, reassociate, the errors clear and the server boots.

I have another Flexpod in an adjacent data center configured the same. Only without this issue.

Thanks,

Hamish

[abbharga]

Hi Hamish,

So just to understand, you only get this fault when you reset the server? If so does the server OS boots up successfully after the reset, if the OS is down then the VIFs can remain down.

What version are you running?

./Abhinav

[HamR]

Hi Abhinav,

I sometimes get this fault when I reset the server. The VIFs are down when the server is down, but they sometimes stay down when the server is booted. After a SP reassociation the VIFs always come up.

The servers are running 2.0(2q) code on B200 M2 blades with ESXi 5.0 U1.

Thanks,

Hamish