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


Re: [Xen-users] Dom0 Crashes on High I/O

To: xen-users@xxxxxxxxxxxxxxxxxxx
Subject: Re: [Xen-users] Dom0 Crashes on High I/O
From: Mark Williamson <mark.williamson@xxxxxxxxxxxx>
Date: Sun, 6 Apr 2008 19:11:59 +0100
Cc: Henri Cook <henri@xxxxxxxxxxxxxxxxxxxxx>, Stefan de Konink <skinkie@xxxxxxxxx>, "Ross S. W. Walker" <rwalker@xxxxxxxxxxxxx>
Delivery-date: Sun, 06 Apr 2008 11:12:36 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
In-reply-to: <13287.>
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/cgi-bin/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=unsubscribe>
References: <13287.>
Sender: xen-users-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: KMail/1.9.6 (enterprise 0.20070907.709405)
> I have mentioned it on xen-devel, and probably on bugzilla. And is one of
> the reasons to use iSCSI over NFS. Someone told me that it could be fixed
> by setting the memory overcommit for Linux to 'do not do it'. That is some
> setting in /etc/sysctl.conf


To help people understand what you're seeing, can you tell us some more about 
your setup?  Is it possible that it's a problem like this (which may be 
provoked by such setups as virtual disk files stored on NFS, etc...).

It'd be useful to know whether you'd been able to try the upstream XenSource 
XenLinux (which really ought to work, if anything does - otherwise it's 
definitely an upstream XenLinux bug!).  Out of interest, where did you source 
the rest of your Xen 3.2 installation - Ubuntu packages, source tarball from 
Xensource, or ... ?

Also, I gather this machine is in a remote hosting facility somewhere?  Is it 
possible that some kind of kernel panic message is being printed to the 
display when crashes?  Can someone in the datacentre have a look?  Better 
than that would probably be to get a serial line hooked up so that any 
crash-time output can be captured - even if the system is too far gone to log 
it to disk.  Do you think this might be possible?


Push Me Pull You - Distributed SCM tool (http://www.cl.cam.ac.uk/~maw48/pmpu/)

Xen-users mailing list

<Prev in Thread] Current Thread [Next in Thread>