The VMware Knowledge Base provides support solutions, error messages and troubleshooting guides
Time in virtual machine drifts due to hardware timer drift (1006072)
- Time in a virtual machine drifts because the hardware time source used by the virtual machine monitor drifts
- Time in a virtual machine drifts at a constant rate
The term time drift describes situations where the reported time diverges from the correct time at an approximately constant rate.
An effective way to measure the drift rate of the hardware being used by ESX is to run NTP in the Service Console and look at the file/var/lib/ntp/drift . NTP uses this file to record the rate of drift that it is correcting.
For example, "-130.824" means that the hardware is running about 130 parts per million (PPM) too fast. Similarly, "45.253" indicates that the hardware clock is running about 45 PPM too slow.
Note: This check only applies to ESX. For other products such a Workstation and Server, the hardware clock used by the host operating system may not correspond to the hardware clock used by the virtual machine monitor to drive the virtual time devices.
All hardware timers run at a slightly different rate than their specified frequency. Operating systems are already designed with this in mind and generally run software to correct for this drift, like NTP or w32time. Because of this, the virtual machine monitor does not attempt to correct the drift of the hardware that it uses for determining the time. This results in the hardware drift being passed on to the guest operating system. On ESX, running NTP in the Service Console corrects time in the Service Console, but does not affect the time used by the virtual machine monitor and the guest operating system. Similarly, on Workstation, Server, Player, Fusion, and ACE, running NTP, w32time or other time synchronization software on the host operating system corrects time on the host, but does not affect the time used by the virtual machine monitor and the guest operating system.
Run time correction software in the guest operating system. Options include NTP, w32time, or VMware Tools periodic time synchronization.
Note: If the time drift is forward, VMware Tools periodic time synchronization does not properly correct it and NTP or w32time must be used.
For related issues, see:
- Severe hardware time drift may be caused by the hardware HPET mis-reporting its frequency. For more information, see Confirming HPET is not misreporting its frequency (1006090) .
- Forward time drift in virtual machines running Linux 2.6 kernels may also be caused by lost tick overcompensation. For more information, see Time in a Linux guest operating system runs faster than real time due to lost tick overcompensation (1006113).
- Forward time drift in virtual machines running Windows may also be caused by problems with the Windows Multimedia timer. For more information, see Time runs too fast in a Windows virtual machine when Multimedia Timer interface is used (1005953).
Request a Product Feature
To request a new product feature or to provide feedback on a VMware product, please visit the Request a Product Feature page.