Dnia poniedziałek, 9 listopada 2009 o 13:32:11 Jordi Espasa Clofent
napisał(a):
> > Jordi,
> >
> > Sorry for writing the above line in a hurry.
>
> No problem.
>
> > I observe, that my domU clock differs from dom0 clock from the very
> > beginning. Thus I think it's something wrong with xen tools that setup a
> > new domain or the prblem is caused by hypervisor or kernel that miss some
> > number of clock ticks during kernel boot.
>
> IMHO this is important. Is there any related bug at "official level"?
>
> > In my combination of Xen(3.4.1)/kernel(2.6.31.5/jeremy) I have no such
> > entry in the /proc tree or I don't know how to make independent_wallclock
> > entry apear somewhere there. But the workaround works without writing to
> > this entry.
>
> Ok.
>
> > Well, it seems that hypervisor's mechanisms of sharing common clock
> > between domains don't work. What works, is ntpd based synchronization
> > independent of xen. Ntpd daemon works in userspace and it behaves better
> > when the system is not fully loaded. The load I was talking about is my
> > favorite load test: linux kernel compilation, do it again and again using
> > at least 20 concurent jobs.
>
> So, I understand the problem could be if you rely in ntpd (userspace) to
> keep up the time sync and the domU is under high load (kernel space). No?
>
> Anyway, the solution is clear: the hypervisor must to do it correctly.
>
Jordi,
I just got answer from Jeremy Fitzhardinge to my questions regarding
timekeepeing sent to xen-devel. See:
http://lists.xensource.com/archives/html/xen-devel/2009-11/msg00505.html .
The good news for me is that it's not that bad as I thought with 2.6.31.5
kernels. Generally, xen makes dom0 and domU clocks go in lockstep with its
wallclock, but at the moment you cannot update xen wallclock from dom0 (nor
from domU). When this feature is implemented, the problem should disappear in
2.6.31+ kernels. Util that moment I'll have to run ntpd in all VMs. I think
200ms time difference, which I have under heavy load, is not a big issue. Of
course, that's my point of view and your applications might be more strict.
If you are interested in older kernels, please, join the discussion on xen-
devel: http://lists.xensource.com/archives/html/xen-
devel/2009-11/msg00469.html
--
Bartosz Lis @ Inst. of Information Technology, Technical Univ. of Lodz, Poland
bartoszl @ ics.p.lodz.pl
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|