Upgrading from ESXi 5.5 to 6.5, performance of the host was relatively poor on a JetWay JNF9G-QM77. Guests were extremely sluggish.
There were also several errors tracked in the various logs related to the boot drive and the datastore associated with the boot drive.
In the vobd log:
[vmfsCorrelator]: [esx.problem.vmfs.heartbeat.timedout] datastore1 [vmfsCorrelator]: [vob.vmfs.heartbeat.recovered] Reclaimed heartbeat for volume (datastore1): [Timeout] [vmfsCorrelator]: [esx.problem.vmfs.heartbeat.recovered] datastore1
In the vmkwarning log:
WARNING: NMP: nmp_DeviceRequestFastDeviceProbe:237: NMP device "" state in doubt; requested fast path state update...
In the vmauthd log:
ScsiDeviceIO ... CmdSN 0x4305 from world 69114 to dev "" failed ... Invalid sense data ...
The solution was to disable the new AHCI driver of ESXi 6.5 (vmw_ahci). From the ESXi console issue the following command:
esxcli system module set --enabled=false --module=vmw_ahci
Then after a reboot ESXi should be using the sata–ahci driver. You can confirm this by checking the list of storage adapters for the host, looking for one titled Panther Point AHCI Controller.
After this driver swap, performance of the host should be restored.