vMotion Compatibility Alert "Too large clock skew was detected" on Time Synced ESXi Hosts

Products

• VMware vSphere ESXi
• VMware vCenter Server

Issue/Introduction

• There is time skew of less than 5 minutes on ESXi hosts, when attempting to run a vMotion job the migration wizard shows the compatibility alert message "Too large clock skew was detected. Relative time skew ######## between source and destination hosts is greater than five minutes."

    NOTE: vMotion jobs still continue to work.

• In ESXi host /var/run/log/vpxa.log you see below entries

        ####-##-##T##:##:##.### In(###) Vpxa[#######]: [Originator@#### sub=vpxaMoService opID=###########################] Time difference between vpxd and vpxa is 412 seconds

Environment

• ESXi 7.x
• ESXi 8.x
• VC  7.x
• VC 8.x 

Cause

A clock skew was previously recorded, but resolved during a time when the VPXA service on the ESXi host was down (For example, a host reboot). When this time skew issue was resolved while VPXA is down, the service is unable to communicate to the vCenter's VPXD service that the skew was resolved. Due to this, the vCenter still believes a time-skew exists.

Resolution

This issue is resolved in vCenter 8.0 Update 3g. Log in to the Broadcom Support Portal to download this patch

To workaround this issue restart the VPXD service using the command below in vCenter

• vmon-cli -r vpxd

       or

• service-control --restart vpxd