This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


[Xen-users] Time gaps in system logs related to NTPD

To: xen-users@xxxxxxxxxxxxxxxxxxx
Subject: [Xen-users] Time gaps in system logs related to NTPD
From: Steven Timm <timm@xxxxxxxx>
Date: Tue, 08 Dec 2009 09:26:34 -0600 (CST)
Delivery-date: Tue, 08 Dec 2009 07:27:25 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
List-help: <mailto:xen-users-request@lists.xensource.com?subject=help>
List-id: Xen user discussion <xen-users.lists.xensource.com>
List-post: <mailto:xen-users@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-users-bounces@xxxxxxxxxxxxxxxxxxx
I am running xen-3.03-94 as bastardized by Redhat and copied by
Centos and Sci. Linux, supposedly it is xen 3.1.2 backported.
This is on the equivalent of RHEL5/update 3.

In the past week I have seen four of my domU's
have the following problem:

It suddenly becomes impossible to access them from the network,
can't ssh in, can't su or ksu from processes that are inside the
node already and running.  About 30-45 minutes later
a message appears in the logs

2009-12-06T09:11:02-06:00 s_sys@xxxxxxxxxxxxxx ntpd[1485]: no servers reachable 2009-12-06T09:19:24-06:00 s_sys@xxxxxxxxxxxxxx ntpd[1485]: synchronized to, stratum 2

What is freaky is that there is nothing at all in any of the logs
during that time.  As far as the OS and all of its processes
are concerned, the time frame between 07:58 and 09:11 didn't exist,
yet there is no entry in the logs saying that ntpd kicked up
a huge time change when it came back on.

Two questions

1) should I be running ntpd at all?  Old advice from a couple of
years ago says no but that was a different kernel branch and
a different everything.  These days ntpd is turned on by
default in redhat's virt-install and I have dozens of other
domU's running the same kernel and ntpd and this configuration
where I don't have this problem.

2) The symptoms are consistent both with a crazy ntpd broadcast
(our ntpd is configured to use a broadcast client) or with
the domU just being hung for that length of time and not
doing anything.  Which is more likely and how do I tell?

There is no evidence that either the domU or dom0 ever lost
network connectivity during this time but i am
seeing about 50K dropped packets on the physical peth0 of the
dom0 and about the same amount on several of the vifn.n, this
for a system which has been up about a month.

Steve Timm

Steven C. Timm, Ph.D  (630) 840-8525
timm@xxxxxxxx  http://home.fnal.gov/~timm/
Fermilab Computing Division, Scientific Computing Facilities,
Grid Facilities Department, FermiGrid Services Group, Assistant Group Leader.

Xen-users mailing list

<Prev in Thread] Current Thread [Next in Thread>
  • [Xen-users] Time gaps in system logs related to NTPD, Steven Timm <=