OV 4.0 issue: interconnect 2 port 10 subport a is unlinked.

[Matthew Ingram]

After an upgrade to OV 4.0 I am seeing an issue with some profiles showing flex nic links in an error state. They show "unlinked" on the interconnects. However, the link are actually up and fine in the guest OS. A reboot of the server doesn't change how its being reported in OV.

An error has occurred on connection 1. Interconnect xxxx, interconnect 2 port 10 subport a is unlinked.

[peyrache]

Try to reapply profile or pull out server and push in againg

[Matthew Ingram]

Niether efuse or reapply of the server profile resolved the issue. It is also not fixed in 4.00.07.

[peyrache]

May be related to this thread
https://community.hpe.com/t5/HPE-OneView/Self-signed-certificate-Error-after-Upgrade-OneView-4-0/td-p/6993311

[ChrisLynch]

The message is originating from the Interconnect, not the Server Profile.  Perform a refresh of just the Interconnect in question or the Logical Enclosure.  Do this during non-peak hours.

[Matthew Ingram]

I agree, but the only change was the OneView upgrade from 3.00.07 to 4.00.05 then to 4.00.07. The error occurs when the server is rebooted. The error occurs during the POST and never clears. For a small number of servers the error clears when the system POST completes. I have a case opened and they recommended trying a VC reset. A reset of the VC clears the error until the server is rebooted again. I have two datacenters each with their own OV appliance. each DC has over 20 C7000 chassis. Almost 50% of my innterconnects are in an error state at this point due to servers being restarted.

[ChrisLynch]

Check the firmware and driver version of the adapter that is mapped to the interconnects.  Can you provide me the model and the versions you have?  If your case is in Level 2, as them to help verify the DCC state of the connection within the Logical Enclosure support dump.

[Matthew Ingram]

We have a mix of BL460c Gen 8, BL660c Gen 8, and BL460c Gen9 servers. The SPP vary from 2017.10.1, 2017.04.0, and June 2015. They are mostly ESXi 5.5 but it also occuring on the latest HPE custom iso of 6.5 with the 2017.10.1 SPP firmware baseline. All of our VC's are currently at 4.45.

[Aaron Phillips_1]

Matthew,

I am having a very similar issue with one of my BL460c Gen9 servers. My subports C and D are giving me the same "unlinked" error. What ended up being your solutiuon?

Thank you,

[ExploitVinci]

Hello.

I have the same issue and the resolution is quite simple.

It is just Oneview which has changed : this alert is linked to the fact that you have deactivated your card in your OS.

For example, I have two network cards in my server profile and I only used one of them. So I have deactivated this card in Windows... No error in Oneview 3, but error in Oneview 4 !!

I have reactivated this network card and the alert is gone !!

Enjoy ^^

[AkiraX]

I just ran into this on a new Synergy build out with SY 480 Gen10, 3530C/4820C. Ended up RESETTING and REFRESHING each interconnect pair(3,6) in each frame. Adapters that were disconnected(windows 2016) now show status UP.

Support on stanby.