virt November 2009

virt@lists.centos.org

44 participants
36 discussions

Fedora 12 2.6.31.5-127.fc12 domU on CentOS 5.4 2.6.18-164.6.1.el5xen fails to boot
by Charles J Gruener 06 Mar '10

06 Mar '10

Doing a kickstart install of Fedora 12 results in a non-bootable image. >From a virt-maanger instance, the error produced when trying to run the domU is: Traceback (most recent call last): File "/usr/share/virt-manager/virtManager/engine.py", line 498, in run_domain vm.startup() File "/usr/share/virt-manager/virtManager/domain.py", line 573, in startup self.vm.create() File "/usr/lib64/python2.4/site-packages/libvirt.py", line 287, in create if ret == -1: raise libvirtError ('virDomainCreate() failed', dom=self) libvirtError: POST operation failed: xend_post: error from xen daemon: (xend.err "Error creating domain: Boot loader didn't return any data!") If you drop to the command line to see what pygrub gets back, you get: # pygrub /var/lib/xen/images/fedora12.img Traceback (most recent call last): File "/usr/bin/pygrub", line 677, in ? fs = fsimage.open(file, get_fs_offset(file)) IOError: [Errno 95] Operation not supported The kickstart worked perfectly in Fedora 11 on a CentOS 5.4 dom0. I've tried investigating all sorts of grub options and even tried copying kernel and initramfs files out from the domU to dom0 and booting from those with no luck. The only thing that does work is if I install Fedora 11 and then perform an upgrade to Fedora 12. I really don't want to have to deploy machines this way. Is anyone else experiencing this issue? I can provide any further information requested. Basically, to repeat my error, install CentOS 5.4 with Virtualization, then try to install a Fedora 12 domU. The installation finishes without error but you can't boot. The systems are all x86_64. Thanks. Charles

9 31

Live migration and DRBD
by Gilberto Nunes 02 Dec '09

02 Dec '09

Hi folks I deploy a two Dell PowerEdge T300 to test Virtualization with kvm+drbd+heartbaet. The KVM drbd and heartbeat work properly. However, I have doubt!! When the primary node has down, the secondary node start the VM that has original running on primary node... So, this required a full stop of hole system... This is not we wish here... Is there something way to live migrate VM from primary node that was shutdown???? I have no idea how to make this stuff working... Thanks for any help -- Gilberto Nunes

7 16

Re: [CentOS-virt] Fedora 12 2.6.31.5-127.fc12 domU on CentOS 5.4 2.6.18-164.6.1.el5xen fails to boot
by Christopher G. Stach II 01 Dec '09

01 Dec '09

----- "Scot P. Floess" <sfloess(a)nc.rr.com> wrote: > It sure does...I am running it now... This has affected F8, F10, 5.3, and so on for a few years. I wouldn't count on it ever working unless you dig up the patches and figure out how to apply them yourself, or find someone who has. :) If you want, you can dig up all of the issues in Red Hat's bugzilla. Here's a recent one that may help you out: https://bugzilla.redhat.com/show_bug.cgi?id=528053 -- Christopher G. Stach II

3 3

Re: [CentOS-virt] Controlling allocation of ethernet devices and KVM
by Christopher G. Stach II 30 Nov '09

30 Nov '09

----- "Steven Ellis" <steven.ellis(a)bulletin.net> wrote: > My issue is I can't see any way to bring up NAT guests unless they are > using a ethernet interface that has a address for the host OS. How will you NAT without an interface address to translate to? You can probably wrangle the scripts into supporting your own configured mangle/NAT setup with something handling the ARP for you on the unnumbered interface, but that's probably beyond the scope of this list. -- Christopher G. Stach II

1 0

Controlling allocation of ethernet devices and KVM
by Steven Ellis 30 Nov '09

30 Nov '09

Running Centos 5.4 with KVM on a Dell R610 server and I'd like to control which of the four ethernet interfaces are used for specific tasks My ideal configuration would be eth0 - Host traffic only, no virtual guests. Used for guest mirroring and management. eth1 - NAT guest traffic only, no address for local machine and in some environments in the same zone as eth0 eth2/3 - Allocated to two different bridge devices which might be in separate network zones. The configuration of eth2/3 is fairly simple, my issue is restricting any NAT traffic to a specific ethernet devices, and ideally one with no local IP. Any ideas? Steve -- *Steven Ellis - Director of Worldwide Engineering,* *Bulletin.Net Inc* - http://www.bulletin.net/

6 15

What kernel source and how to download it
by Mauro Gatti 27 Nov '09

27 Nov '09

Hi all, I am running CentOS 5.3 under xen on a VPS machine (so I have a limited control on the machine) I am playing with openswan and KLIPS and I need to build the ipsec.ko kernel module. I would need to download kernel source but I am really confused about what source I have to get. uname -r tells: 2.6.18-128.7.1.el5xen ls /lib/modules tells: ls /lib/modules 2.6.18-128.7.1.el5xen 2.6.18-164.2.1.el5 2.6.18-164.6.1.el5 2.6.18-53.1.13.el5xen I have no sources in /usr/src/kernels Is there a source code named "2.6.18-128.7.1.el5xen" which I can download or should I install kernel-devel source using yum? If I used yum I would download 2.6.30.9... yum info kernel-devel Loaded plugins: fastestmirror Loading mirror speeds from cached hostfile Available Packages Name : kernel-devel Arch : i686 Version : 2.6.18 Release : 164.6.1.el5 Size : 5.2 M Repo : updates Summary : Development package for building kernel modules to match the kernel. URL : http://www.kernel.org/ License : GPLv2 Description: This package provides kernel headers and makefiles sufficient to build modules : against the kernel package. and kernel.org has only source code from 2.6.18.1 to 2.6.18.8 Thanks Mauro

3 2

Virsh shutdown all command?
by Neil Aggarwal 25 Nov '09

25 Nov '09

Hello: Is there a command in virsh to shutdown all domains? I can do one at a time, but that is untenable for a large number of domains. Thanks, Neil -- Neil Aggarwal, (281)846-8957, http://UnmeteredVPS.net CentOS 5.4 VPS with unmetered bandwidth only $25/month! 7 day no risk trial, Google Checkout accepted

4 6

Shell command to list cpu usage for KVM guests?
by Neil Aggarwal 25 Nov '09

25 Nov '09

Hello: I am looking for a way to check CPU usage for KVM guests. I see that capability in virt-manager, but that would require me to install a GUI on the host. Is there a simple report from a command line tool? Something like the three load average numbers per guest in a list? I took a look at virsh but could not find anything. Thanks, Neil -- Neil Aggarwal, (281)846-8957, http://UnmeteredVPS.net Host Joomla!, Wordpress, phpBB, or vBulletin for $25/mo Unmetered bandwidth = no overage charges, 7 day free trial

2 1

dom0 hangs when rebooting debian lenny domU
by Günter Zimmermann 25 Nov '09

25 Nov '09

Dear all, I have a reproducible issue with centos54 dom0 and debian lenny domU using pygrub. As far as I enter "shutdown -r now" in the running Debian Lenny DomU the whole system hangs. Keyboard is dead, network is dead and no more logging. Ending up in a hardware reset of the box. I was able to reproduce this 3 times. Restarting with xm reboot debian50 works without problems. Is there a possibility to debug this behavior? Unfortunately I neither had any logging in /var/log/* nor in /var/log/xen/* or on console. The debian lenny installation was done using debootsrap. No additional software has been installed in the debian domain. This is my cfg file for the Debian Lenny domain: bootloader = "/usr/bin/pygrub" memory = 1024 extra = "3" name = "debian50" vcpus = 1 vif = [ 'bridge=xenbr1' ] disk = [ 'phy:store/debian50_boot,xvda,w', 'phy:store/debian50_root,xvdb,w', 'phy:store/debian50_usr,xvdc,w', 'phy:store/debian50_var,xvdd,w', 'phy:store/debian50_tmp,xvde,w', 'phy:store/debian50_home,xvdf,w', 'phy:store/debian50_swap,xvdg,w', 'phy:store/compile,xvdh,w'] hostname = "debian50.zimmermann.com" on_poweroff = 'destroy' on_reboot = 'restart' on_crash = 'restart' pci = [ '0000:09:04.0' ] Best Regards, Günter Zimmermann

1 0

Unknown boot option `pciback.hide=(09:04.0)': ignoring
by Günter Zimmermann 25 Nov '09

25 Nov '09

Good morning, I need to use pciback.hide to hide a pci card because I want to pci passthrough that card to a domU. My hardware neither supports Intel-Vt-d nor AMD's IOMMU. Unfortunately this doesn't seem to work in Centos 54 anymore. I guess it doesn't work because it is not compiled into the kernel, but as a module. I got the following error in dmesg and the pci device is still listed in dom0 in lspci. Bootdata ok (command line is ro root=/dev/dom0/root pciback.hide=(09:04.0)) Kernel command line: ro root=/dev/dom0/root pciback.hide=(09:04.0) Unknown boot option `pciback.hide=(09:04.0)': ignoring I also tried to update initrd to include the module but got the same error message as above. mkinitrd --with=pciback /boot/initrd-2.6.18-164.6.1.el5xen_pcipt 2.6.18-164.6.1.el5xen title CentOS (2.6.18-164.6.1.el5xen PCI PT) root (hd1,0) kernel /xen.gz-2.6.18-164.6.1.el5 module /vmlinuz-2.6.18-164.6.1.el5xen ro root=/dev/dom0/root pciback.hide=(09:04.0) module /boot/initrd-2.6.18-164.6.1.el5xen_pcipt Is there any option to use pciback.hide kernel parameter without compiling a custom kernel? Thanks in advance Günter Zimmermann

2 3

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

virt November 2009