> From: Frank Mehnert [mailto:[email protected]]
> 
> to a vanilla kernel anymore. They are completely independent kernels. But
> newer kernels seem to be able to better calibrate their timer in
> virtualized environments. Unfortunately they use a guest clock frequency
of
> 1000Hz which makes it even harder to satisfy their timing demands.
> ...
> As written above, there is is no general fix for this problem. Decrease
the
> concurrent load on the host. Use the divider=10 parameter for the guest
> kernel. This will force the guest kernel to use a clock frequency of 100Hz
> which should help a lot. Try to use only one guest CPUs.

Thanks - divider=10 seems to have fixed the problem.  But it just seems so
arbitrary and unscientific.

If I understand correctly, the kernel is frequently adjusting the system
time according to a timed schedule, but the virtualbox addons are also
adjusting the system clock.  Unfortunately, the kernel wasn't able to
correctly calibrate itself, and the kernel is running the updates so darn
fast, virtualbox can't keep up, an virtualbox simply gets overwhelmed.  So
by dividing the kernel by 10, we haven't *actually* solved the problem, but
we've lowered the kernel's influence and thus allowed virtualbox to become
the winning factor.  Is that right?

Now that I have divider=10, I see the system time fluctuate, fast and slow,
as much as 10-20 seconds off the correct time.  But it always comes back,
and it never gets more than a number of seconds wrong.  (It never becomes
wrong by as much as a minute.)

I've been monitoring it while idle, and while under load.  If there is a
behavior condition that will cause this solution to become unstable again, I
haven't been able to identify that behavior.  As far as I can tell, it's
stable now and good enough for me to call it "fixed."



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
VBox-users-community mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/vbox-users-community
_______________________________________________
Unsubscribe:  
mailto:[email protected]?subject=unsubscribe

Reply via email to