From mboxrd@z Thu Jan 1 00:00:00 1970 From: Cris Daniluk Subject: Re: Dom0 panic on xen 3.4.3 Date: Sat, 21 Aug 2010 18:27:41 -0400 Message-ID: References: <20100820235913.GU2804@reaktio.net> <20100821101223.GV2804@reaktio.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: <20100821101223.GV2804@reaktio.net> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: =?ISO-8859-1?Q?Pasi_K=E4rkk=E4inen?= Cc: xen-devel@lists.xensource.com List-Id: xen-devel@lists.xenproject.org On Sat, Aug 21, 2010 at 6:12 AM, Pasi K=E4rkk=E4inen wrote: > On Fri, Aug 20, 2010 at 08:20:54PM -0400, Cris Daniluk wrote: >> On Fri, Aug 20, 2010 at 7:59 PM, Pasi K=E4rkk=E4inen wrot= e: >> > On Fri, Aug 20, 2010 at 02:15:25PM -0400, Cris Daniluk wrote: >> >> I'm running Xen 3.4.3 final and observing the following kernel panic >> >> when I fire up the 9th domU. dom0 is CentOS 5.5 running a XenLinux >> >> kernel. grub.conf entry is below, followed by the dump. The kernel >> >> crash is immediately after the 9th domU is unpaused. >> >> >> >> title Xen 3.4.3 CentOS (2.6.18-194.11.1.el5xen) >> >> =A0 =A0 =A0 =A0 root (hd0,0) >> >> =A0 =A0 =A0 =A0 kernel /xen-3.4.3.gz dom0_mem=3D768M loglvl=3Dall gue= st_loglvl=3Dall >> >> com1=3D38400 >> >> 0,8n1 console=3Dcom1 sync_console noreboot >> >> =A0 =A0 =A0 =A0 module /vmlinuz-2.6.18-194.11.1.el5xen ro root=3D/dev= /vg_internal/lv_base >> >> console=3Dhvc0,38400 norhgb netloop.nloopbacks=3D100 >> >> =A0 =A0 =A0 =A0 module /mpp-2.6.18-194.11.1.el5xen.img >> >> title CentOS (2.6.18-194.11.1.el5) with MPP >> >> =A0 =A0 =A0 =A0 root (hd0,0) >> >> >> > >> > so I assume this crash doesn't happen if you use the el5 default xen h= ypervisor? >> > >> > Have you tried using latest linux-2.6.18-xen (from xen.org) ? >> > >> > -- Pasi >> > >> >> The dom0 kernel is the el5 default, but I have not tried running their >> hypervisor since it is so dated. I'll give a shot with the latest >> xen.org kernel. >> > > Well the version number looks old (3.1.2), but it has a lot of fixes and = backports > from newer xen versions. > > -- Pasi > > Having the same issue with the EL5 RPM as well. Latest 2.6.18 on xen.org seems to be a little broken. The LSI controller isn't getting detected after the megaraid_sas module loads. The initrds look identical between the two, so I'm assuming there is some obscure bug that was fixed and backported into the EL5 kernel but not Jeremy's. I suppose I can try a newer pvops kernel if you think that might be relevant. Seems to be fairly consistently crashing regardless of Xen version, though. Also interestingly it seems it may be more about the VMs starting at once htan the specific number of VMs.. Cris