Locked alert will not clear

[JSmid]

I have one enclosure in OneView that is having some problems. At some point about two months ago OV lost connection to the ILO of the blades and threw an alert, this alert will not clear.

"Network connectivity has ben lost for server hardware *****, bay 9. Resolution: Check and restore network connectivity to the management processor. Refresh the server."

Since this alert appeared,I have tried several things to clear this alert but I cannot get it to clear. Things i have tried.

Reset the ILO

Reset the blade

Refresh the enclosure

Refresh the hardware

Re-Apply the HW profile

Change a setting in the profile

Update the firmware on the blade and the enclosure

Everything works, but i cannot get this alert to clear. Any help would be appreciated.

In the meantime, i've also updated from 1.20.5 to 2.0 which did not solve it.

[ChrisLynchHPE]

When an Alert is Locked, it is because the physical condition has not been fixed or resolved.  Does the OA report any communication loss with the iLO?  Can you log into the iLO remotely via SSH or the Web Admin Console?

[JSmid]

The OA does not report any problems accessing the ILO. I can also log into the iLO via SSH or wia the Web interface. From Oneview, I can also refresh the hardware and reboot it. The refresh finishes correctly, however the server hardware still shows the Error Alert. What I have done to test this is to remove some settings frm the iLO web interface such as the user account and the NTP settings, when I run a refresh it correctly sets these settings.

[ChrisLynchHPE]

Hmm...  Is the hypervisor hosting the OneView appliance being over-taxed? Is there enough disk (IOPs, not capacity) and CPU resources?

[JSmid]

Yes, there are more than enough resources available. There's not much else running on this cluster.

[JSmid · Accepted solution]

@MeFromil posted the solution in the following topic:

http://community.hpe.com/t5/HPE-OneView/Clearing-locked-alert-status/m-p/6815799#M908

[Afridi43]

Kindly perform a clear NVram on the server with the profile locked.

 1-     Power off the server and locate the system maintenance switch  ( check the mapping on server’s access panel)

2-     Move switch S6 from default 'Off' position to 'On'

3-     When a message indicates that a maintenance switch has been turned on and the unit has been reset, power it off

4-     Reverse switch S6 and reboot again

[Gary Mueller]

I am having the same issue but with a gen9 dl380. This happened when I applied a profile to this server