From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Subject: Re: Loop Back Mounted Files Cannot be UMounted, DomU Cannot Restart Date: Wed, 14 Mar 2007 14:28:26 +1300 Message-ID: <45F74FBA.5090301@rimuhosting.com> References: <45F5E5A2.6040704@rimuhosting.com> <20070313100540.GE30259@leeni.uk.xensource.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20070313100540.GE30259@leeni.uk.xensource.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Ewan Mellor Cc: xen-devel@lists.xensource.com List-Id: xen-devel@lists.xenproject.org Ewan Mellor wrote: >> [2007-03-08 21:53:48 xend 19559] DEBUG (XendDomain:198) Cannot recreate >> information for dying domain 163. Xend will ignore this domain from now on. > > This looks suspicious too. It looks like you have a VM that has not been > properly cleaned up -- presumably it's the one that's holding your loopback > device open. Do you have the logs for what happened to domain 163? > > Ewan. > Did you have any particular logs in mind? I think it was likely just from an xm destroy. Though an xm destroy does not always lead to this problem. Anyway to recover from this? Force Xen to release the loop back mounted file:/ device? Recap on our environment: Centos4, hotplug, loop back mounted files, about 60 odd domUs, "page allocation failure" errors (typically with networking code in the stack dump), xen 3.0.3, sometimes it seems like the problem is triggered when there is high disk io load. Any ideas or suggestions appreciated. Regards, Peter