I recently upgraded our demo lab to vSphere 6.7, despite the known issues with VEEAM and other tools not being ready for the big changes 6.7 brings with it. I was curious to see the new web interface and the other new features and well, what is a demo lab for if not for testing....
VCSA upgrade was easy from 6.5 to 6.7 (except you have made the mistake and installed 6.5 U2 on any of your systems, then upgrade is not supported at all) and a few hours later I upgraded the first of our three HPE ProLiant systems to ESXi 6.7. The frst host is a DL380p Gen8 system with a single CPU. Running on top of vSphere 6.5 I never saw a hardware status error but immediately after upgrading to 6.7, vCenter shows a hardware failure with the system fans.
VMware obviously still has a problem with translation of error messages already created so the text is still in german.... Don't care, it only says that the fan status is critical.
Curious, a few hours before under 6.5 everything was fine and now one or more of the fans should have died? I checked the iLO to get more information about the hardware.
Everything seems to be fine but what causes the error in vCenter? The answer is quite simple. If you click on the "show empty bays" link you will see that there is fan 1&2 unpopulated due to the fact that this server only has 1 CPU.
Unfortunately iLO reports these fans not as "not installed" the the vCenter agents but it reports the current speed of "0"%. This is interpreted by vCenter as a failure as a fan should never run at 0% speed and that's what causes vCenter to raise the alert.
The positive side is that is is a false positive and everything is okay. The negative side is, you can't suppress or disable the error in vCenter. If you rely on hardware information through vCenter, all your single CPU hosts will show up as "critical" and that's disturbing and annoying.
VMware is aware of that bug and will probably release a patch within the next weeks. Until then, simply ignore these kind of errors (if you can).