virt

virt@lists.centos.org

2 participants
1613 discussions

Re: [CentOS-virt] Migrating from KVM to XEN - kernel panic
by Christopher G. Stach II 09 Nov '09

09 Nov '09

----- "Christopher G. Stach II" <cgs(a)ldsys.net> wrote: > 1. mkinitrd --with=xenblk --with=xennet --preload=xenblk > --preload=xennet <img> <version> > (yes, that's the original one without the fstab) > 2. gunzip -c /boot/initrd-blah.img > /root/blah > 3. mkdir /mnt/blah > 4. mount -o loop /root/blah /mnt/blah > 5. vi /mnt/blah/init > <change occurrences of dom0 VG to guest VG> > <change dom0 root device name to guest root device name> > 6. umount /mnt/blah > 7. gzip -c /root/blah > /boot/initrd-blah.img If you followed these directions, you're probably scratching your head. I screwed up. The initrd is actually a cpio archive, so: 1. mkinitrd --with=xenblk --with=xennet --preload=xenblk --preload=xennet <img> <version> (yes, that's the original one without the fstab) 2. mkdir /root/initrd-blah 3. cd /root/initrd-blah 4. gunzip -c /boot/initrd-blah.img | cpio -idmv 5. vi /root/initrd-blah/init 6. find . -print | cpio -o | gzip -c > /boot/initrd-blah.img -- Christopher G. Stach II

2 1

Re: [CentOS-virt] Migrating from KVM to XEN - kernel panic
by Christopher G. Stach II 09 Nov '09

09 Nov '09

----- "Christopher Hunt" <dharmachris(a)gmail.com> wrote: > I've updated accordingly, and now am getting a brand new file system > issue: > > > Waiting for driver initialization. > Scanning and configuring dmraid supported devices > Creating root device. > Mounting root filesystem. > mount: could not find filesystem '/dev/root' > Setting up other filesystems. > Setting up new root fs > setuproot: moving /dev failed: No such file or directory > no fstab.sys, mounting internal defaults > setuproot: error mounting /proc: No such file or directory > setuproot: error mounting /sys: No such file or directory > Switching to new root and running init. > unmounting old /dev > unmounting old /proc > unmounting old /sys > switchroot: mount failed: No such file or directory > Kernel panic - not syncing: Attempted to kill init! This could be a couple of things. Is it still displaying the message that it found vg00? If not, (1) it's mkinitrd being stupid and not creating the initrd init script correctly (ignoring LVM initialization when you supply an alternate fstab). If it is starting vg00 properly, it is probably (2) that your root device names are different in dom0 and the guest.' If you have situation 1, you can modify the initrd manually or do a lengthy process with kpartx, yum install kernel-xen, bind mount a bunch of filesystems from dom0, mkinitrd, undo everything, and proceed to the next guest. (It's scriptable, bit still a pain.) Modifying the initrd manually isn't that bad. It basically goes like this: 1. mkinitrd --with=xenblk --with=xennet --preload=xenblk --preload=xennet <img> <version> (yes, that's the original one without the fstab) 2. gunzip -c /boot/initrd-blah.img > /root/blah 3. mkdir /mnt/blah 4. mount -o loop /root/blah /mnt/blah 5. vi /mnt/blah/init <change occurrences of dom0 VG to guest VG> <change dom0 root device name to guest root device name> 6. umount /mnt/blah 7. gzip -c /root/blah > /boot/initrd-blah.img You can keep repeat 4-7 as necessary until it works or you lose it. :) If you don't have situation 1 but you have situation 2, just change the root device name in fstab-domU and rerun the mkinitrd command with the fstab argument. -- Christopher G. Stach II

1 0

Re: [CentOS-virt] Migrating from KVM to XEN - kernel panic
by Christopher G. Stach II 09 Nov '09

09 Nov '09

----- "Christopher Hunt" <dharmachris(a)gmail.com> wrote: > Found volume group "vg00" using metadata type lvm2 ## this is the VG > on the KVM host > Volume group "virt01vg00" not found ## this is the VG on the XEN host mkinitrd is getting this from your dom0's /etc/fstab. sed 's/virt01vg00/vg00/g' /etc/fstab > /tmp/fstab-domU mkinitrd -f --fstab=/tmp/fstab-domU --with=xenblk --with=xennet --preload=xenblk --preload=xennet /boot/initrd-2.6.18-164.6.1.el5xen-domU.img 2.6.18-164.6.1.el5xen Better luck this time. :) -- Christopher G. Stach II

2 1

Re: [CentOS-virt] Migrating from KVM to XEN - kernel panic
by Christopher G. Stach II 09 Nov '09

09 Nov '09

----- "Christopher Hunt" <dharmachris(a)gmail.com> wrote: > This brings me back to suspecting the problem is in the different file > structures. I'm still trying to wrap my head around the other replies > in this thread. mkinitrd in dom0 with xenblk and xennet and use that new initrd for the guests. After you get the guests up and running, install kernel-xen in each one and switch to using pygrub. The following should handle the first part, but no warranties implied. :) mkinitrd --with=xenblk --with=xennet --preload=xenblk --preload=xennet /boot/initrd-2.6.18-164.6.1.el5xen-domU.img 2.6.18-164.6.1.el5xen for i in /etc/xen/guest[0-9]*; do cp ${i} ${i}.bak sed -i '/^ramdisk/cramdisk = "/boot/initrd-2.6.18-164.6.1.el5xen-domU.img"' ${i} done -- Christopher G. Stach II

2 1

Domain0 with Kernel 2.6.18-92.1.22.el5xen fails to boot Guests with Kernel 2.6.18-164.el5xen
by Devraj Mukherjee 09 Nov '09

09 Nov '09

Hi all, I am in a bit of a fix here, I have several Guests paravirtualized running on a Dom0 that is currently running the 2.6.18-92.1.22.el5xen Kernel. All my guests (administered by different people) have subsequently been updated to later versions of the kernel (namely 2.6.18-164.el5xen or 2.6.18-128.4.1.el5xen). Ones that are still running are using the Kernel 2.6.18-128.4.1.el5xen while Dom0 is still booted up into 2.6.18-92.1.22.el5xen I restarted a Guest that's now trying to boot with the 2.6.18-164.el5xen Kernel and it spits out messages like XENBUS: Waiting for devices to initialise: 295s...290s... and eventually reboots. If I try and interrupt the boot and choose a previous version of the Kernel on the Guest, it says the following and fails to boot. The guests use a single disk image that has the SWAP and Root partitions. Is this happening because the version of the kernel on the Dom0 is way too old for guests to boot? I am just a bit hesitant in rebooting my server for the new kernel to take effect unless I am sure :) device-mapper: dm-raid45: initialized v0.2429 Scanning and configuring dmraid supported devices Scanning logical volumes Reading all physical volumes. This may take a while... Activating logical volumes Volume group "VolGroup00" not found Creating root device. Mounting root filesystem. mount: could not find filesystem '/dev/root' Setting up other filesystems. Setting up new root fs setuproot: moving /dev failed: No such file or directory no fstab.sys, mounting internal defaults setuproot: error mounting /proc: No such file or directory setuproot: error mounting /sys: No such file or directory Switching to new root and running init. unmounting old /dev unmounting old /proc unmounting old /sys switchroot: mount failed: No such file or directory Kernel panic - not syncing: Attempted to kill init! -- "The secret impresses no-one, the trick you use it for is everything" - Alfred Borden (The Prestiege)

3 4

Re: [CentOS-virt] Question - Xen host, DHCPD and guest VMs
by Christopher G. Stach II 06 Nov '09

06 Nov '09

----- "Scot P. Floess" <sfloess(a)nc.rr.com> wrote: > Anybody know what I need to do :) I've been bouncing dhcpd in > /etc/rc.local and starting any VMs there. Although annoying, I'd > rather > do it "right" so I can auto start my Xen guests... If you want to do it correctly, don't run dhcpd in domain 0. :) You're exposing your host to traffic that it doesn't need to deal with, which can be bad for performance and is bad for security. You're also putting one extra service there that isn't solely related to keeping the ground beneath the guests, the only thing domain 0 should be doing. Start a guest to run dhcpd in auto and name the sym link in auto something like "00-guestname" so it runs before the other guests. If you have two or more physical hosts and you run ISC dhcpd, run another one of these guests as a dhcp failover peer on another host. -- Christopher G. Stach II

4 4

Question - Xen host, DHCPD and guest VMs
by Scot P. Floess 06 Nov '09

06 Nov '09

I have a question about a Xen host that runs dhcpd and installing guest/running guest VMs on that host (that use DHCP)... On other hosts or hosts running VMs, they are absolutely able to get DHCP addresses no trouble... However, the fun begins when the machine I run dhcpd on is also hosting VMs. I can see in /var/log/messages the request and ack from dhcpd but the guest VMs never hear it. If I should bounce dhcpd prior to a guest install or start a guest VM - everything works just fine. I'm guessing, perhaps, there is some iptables rule I need to set up? Based on the init scripts, dhcpd definitely starts before xend and xendomains and ultimately the xen bridge... Anybody know what I need to do :) I've been bouncing dhcpd in /etc/rc.local and starting any VMs there. Although annoying, I'd rather do it "right" so I can auto start my Xen guests... Thanks ahead of time! Scot P. Floess 27 Lake Royale Louisburg, NC 27549 252-478-8087 (Home) 919-890-8117 (Work) Chief Architect JPlate http://sourceforge.net/projects/jplate Chief Architect JavaPIM http://sourceforge.net/projects/javapim Architect Keros http://sourceforge.net/projects/keros

2 3

Do not see domain0 on virt-manager...
by Gilberto Nunes 05 Nov '09

05 Nov '09

Hi folks... I see a think realy stranger here... I have two node with centos 5.4 and running xen on both nodes. On node-0 when i run virt-manager, I do not see Domain0 but when I run virsh list, I can see it On node-1, i can see the domain0 on virt-manager... What is wrong! Thanks -- Gilberto Nunes

2 1

Windows Server 2008 x86_64 guest with XEN on CentOS 5.3
by Sergio Cioban Filho 04 Nov '09

04 Nov '09

Hi all, I'm trying to run Windows Server 2008 x86_64 on XEN, but I have a problem... The instalation works fine, but at the first reboot after instalation I get an "Stop 0x0000007E" error. Please, can you help me with this problem? My *processor*: Intel(R) Xeon(R) CPU E5410 @ 2.33GHz My *xm info*: xm info host : vs01.lan.specto.com.br release : 2.6.18-164.2.1.el5.plusxen version : #1 SMP Fri Oct 9 13:06:25 EDT 2009 machine : x86_64 nr_cpus : 4 nr_nodes : 1 sockets_per_node : 1 cores_per_socket : 4 threads_per_core : 1 cpu_mhz : 2327 hw_caps : bfebfbff:20100800:00000000:00000140:040ce3bd:00000000:00000001 total_memory : 8186 free_memory : 2075 node_to_cpu : node0:0-3 xen_major : 3 xen_minor : 1 xen_extra : .2-164.2.1.el5. xen_caps : xen-3.0-x86_64 xen-3.0-x86_32p hvm-3.0-x86_32 hvm-3.0-x86_32p hvm-3.0-x86_64 xen_pagesize : 4096 platform_params : virt_start=0xffff800000000000 xen_changeset : unavailable cc_compiler : gcc version 4.1.2 20080704 (Red Hat 4.1.2-46) cc_compile_by : mockbuild cc_compile_domain : centos.org cc_compile_date : Fri Oct 9 12:19:50 EDT 2009 xend_config_format : 2 **** Sorry for my poor english :) **** Thanks, Regards, --- Sérgio Cioban Filho - LPIC1 Tecnólogo em Gestão de TI Consultoria, projetos e soluções OpenSource ----------------------------------------------------------------- | Linux :: redes :: servidores :: segurança :: virtualização | cluster :: openldap :: voip :: ShellScript | http://cioban.googlepages.com | +55 48 9989-8733 ----------------------------------------------------------------- ..:: Seja livre, use LiNuX!! ::.. ----------------------------------------------------------------- SOLISC - Não perca o maior evento de Software Livre de Santa Catarina - http://www.solisc.org.br

1 0

DomU second /29 on eth0
by Tracy Phillips 03 Nov '09

03 Nov '09

Hi, I have a question that relates to adding additional IP's to a Xen guest. Do I need to do anything at the Dom0 level to add a /29 to a DomU that has an IP from Dom0's /29? I certainly hope that made sense. Thanks in advance for your help. Tracy Phillips

2 1

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

virt