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

Re: [Xen-users] Help: Guest OS time synchronization problem


  • To: "Chu, Sam" <sam.chu@xxxxxxxxx>
  • From: Nico Kadel-Garcia <nkadel@xxxxxxxxx>
  • Date: Mon, 11 Dec 2006 14:50:58 +0000
  • Cc: xen-users@xxxxxxxxxxxxxxxxxxx
  • Delivery-date: Mon, 11 Dec 2006 06:51:03 -0800
  • Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:user-agent:mime-version:to:cc:subject:references:in-reply-to:content-type:content-transfer-encoding; b=To14MVpLGu2uloGeWdxNuLnKy1TJ3SBYUX9bSa6ox7/ykwoSDekB32pilnH/555ABzYuKJaak1UYraxKSXc46LsNZ9hBLIkSxBDQqI1zaRLXE7CM+cb37IXP6uiEa9z9l0+ZOcfXdH8ZT+BrFQIML2fiqXY+4H4DTvLmYMaR00o=
  • List-id: Xen user discussion <xen-users.lists.xensource.com>

Chu, Sam wrote:

I am having great difficulty to maintain guest OS time. Here is my environment.

Hardware: Intel Woodcrest processor.

Xen: 3.0.3

Xen OS: Red Hat Enterprise 4 Update 3, x86_64 version

Guest OS: configuration is pretty standard. Please see the end of email.

Guest OS: Red Hat Enterprise 4 Update 3, x86_64, with everything installed, ntpd not running.

/proc/sys/xen/independent_wallclock = 0

/proc/sys/xen/permitted_clock_jitter = 10000000


NTP should run out of Dom0: that seems to correctly set NTP for all DomU's.

Do not run independent, local hardware clocks, if you can reasonably avoid it. Seriously: I'm afraid the kind of time drift you're seeing is typical for overclocked, overheated, low quality motherboards with clock chips purchased at an "OEM discount" from the back of somebody's bicycle in Taiwan. Unfortunately, I've seen too much of that kind of hardware in use for Beowulf clusters.


_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.