Re: questions about kvm and accurate clocks

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, Apr 09, 2018 at 11:00:06AM -0600, Chris Friesen wrote:
> Hi,

Hi Chris,

> I was looking at the patch series at
> "https://lkml.org/lkml/2017/1/24/551"; and I had a few questions.
> 
> First, did this or something like it ever get merged?

Yes it was. Instructions to setup can be found at:

https://opensource.com/article/17/6/timekeeping-linux-vms

(At the end of the article).

> Second, is something like this always needed with KVM, or does it
> depend on how the guest keeps track of time?

Only to keep very accurate time between guest and host (depends
on the load of the system), but below 10us in a synthetic test-case.
In case, for example, the host is PTP synchronized via network.

For most use-cases NTP synchronization via network is sufficient.

> From a purely theoretical point of view it seems like it should be
> possible to adjust the time reported by the virtual/paravirtual
> clock sources such that the guest would always see the "correct"
> time as long as the host is keeping "correct" time.
> 
> Thanks,
> Chris Friesen

This KVM PTP driver keeps the difference below the host clock 
and guest clock maxed at 10us. 




[Index of Archives]     [KVM ARM]     [KVM ia64]     [KVM ppc]     [Virtualization Tools]     [Spice Development]     [Libvirt]     [Libvirt Users]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite Questions]     [Linux Kernel]     [Linux SCSI]     [XFree86]

  Powered by Linux