Known issue that shows up when Windows Server 2003 is virtualized under Microsoft Hyper-V when a multi-core CPU is assigned to it

We have received reports from customers about PING sensors that show very high (ten thousand milliseconds or more) or even negative request times. The reason for this behavior can be a known issue of Hyper-V that shows up when Windows Server 2003 is virtualized under Microsoft Hyper-V when a multi-core CPU is assigned to it.

The reason for this behavior can be a known issue that shows up when Windows Server 2003 is virtualized under Microsoft Hyper-V when a multi-core CPU is assigned to it. Hyper-V only guarantees “correct timings” for high resolution timers (which we use to measure PING times in PRTG) for single-CPU virtual machines. When using multiple processors, they can actually have slightly different timers. As such, if an application relies on these timers, measured time differences can actually result in negative or astronomical values, seeing as the timers themselves have no fixed common base.

Possible Solutions:

• Microsoft recommends to use the “/usepmtimer” flag in the boot.ini of the respective guest OS in order to avoid these issues and ensure a correct time measurement.
• Assign a single core CPU to the virtual machine
• Migrate to VMware, et al.

For further reference please see http://blogs.msdn.com/tvoellm/archive/2008/06/05/negative-ping-times-in-windows-vm-s-whats-up.aspx

See Also

• Why do I get negative Ping values on my AMD system?
• Where do I configure the active flow timeout's in PRTG Network Monitor?