Does anyone have a handle on the actual time difference (if any) between the Mac OS time and the Windows Guest time (under XP and 2000)? I am going to be time-tagging events from both OS's simultaneously and measure differences between the data streams in both OS's. I need to know how accurate and the order of magnitude of any jitter between the two clocks, since I am trying to establish deltas in the 1-50 millisecond range between the measurements and I don't want to be introducing additional deltas due to the host/guest time synchronization. Thanks in advance, MNL
Hello, emulation and virtualization technology do not allow to achieve the same time accuracy as the real computer has, there is always a slight error - time is a bit slow. So that is why the Time Synchronization feature was created.
Time Difference There seems to be a fairly steady~0.6 second offset between the VM and the host. I would have thought that this much of a difference could cause issues with various time-stamping tools including database management and syncronization -- but I guess not. It's not clear whether the synchronization via the Parallels Tools is better or worse than running NTP separately on the Host and VM operating systems. Any comments on NTP vs Parallels Tools Time Synchronization would be appreciated.
.6 second offset will usually not interfere even with Kerberos auth. Yes, I am sure if you require a 1 sec. response-time of the system, there is a chance if the delta is large since it is higher than a .5 offset it could be an issue, but usually, note usually, this is not an issue. Using the Parallels sync versus NTP on different machines assures that both the VM and the Host are synched to the same NTP host, as I'm sure you know. If you were to point both the guest and host OS at the same NTP host, I would imagine that it would be roughly the same.
