I finally decided to install the CR repo on one of my CentOS 6 machines that I use as a host for some VMs (also running CentOS 6).
Before updating the host I updated a VM that was not critical to test the process and was able to boot that VM, so I went ahead with updating the host and the two VMs that were in use.
When I rebooted the host none of the VM would start.
Once I figured out that it had to be the host I started downgrading the virtual packages. I was finally able to get the VMs to boot after dwongrading qemu-kvm.
Has anyone else seen this?
Emmett
On Mon, Nov 28, 2011 at 10:38 AM, Emmett Culley emmett@webengineer.com wrote:
I finally decided to install the CR repo on one of my CentOS 6 machines that I use as a host for some VMs (also running CentOS 6).
Before updating the host I updated a VM that was not critical to test the process and was able to boot that VM, so I went ahead with updating the host and the two VMs that were in use.
When I rebooted the host none of the VM would start.
Once I figured out that it had to be the host I started downgrading the virtual packages. I was finally able to get the VMs to boot after dwongrading qemu-kvm.
Has anyone else seen this?
You might want to provide more detailed info. What is the version of qemu-kvm that worked/failed ? What is the kernel version? When the VM guests failed to start, did you see any error message?
Also, VM-related questions may be better discussed on the cent-virt mailing list.
Akemi
On 11/28/2011 02:48 PM, Akemi Yagi wrote:
On Mon, Nov 28, 2011 at 10:38 AM, Emmett Culleyemmett@webengineer.com wrote:
I finally decided to install the CR repo on one of my CentOS 6 machines that I use as a host for some VMs (also running CentOS 6).
Before updating the host I updated a VM that was not critical to test the process and was able to boot that VM, so I went ahead with updating the host and the two VMs that were in use.
When I rebooted the host none of the VM would start.
Once I figured out that it had to be the host I started downgrading the virtual packages. I was finally able to get the VMs to boot after dwongrading qemu-kvm.
Has anyone else seen this?
You might want to provide more detailed info. What is the version of qemu-kvm that worked/failed ? What is the kernel version? When the VM guests failed to start, did you see any error message?
Also, VM-related questions may be better discussed on the cent-virt mailing list.
Akemi _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
kernel-2.6.32-131.17.1.el6.x86_64 (after update and currently)
After update:
qemu-kvm-0.12.1.2-2.160.el6_1.8.x86_64
After downgrade:
qemu-kvm-0.12.1.2-2.113.el6_0.8.x86_64
Prior to installing the CR repo the host was up to date with the CentOS base and updates repos (not many packages different from the release ISO). After, it was up to date with the CR repo and the VM would not start. The host is now up to date with the CR repo except that libvirt* pyton-libvirt and qemu-kvm have all been downgraded to what was up to date for the CentOS base and updates repos.
By "won't boot" I mean that "running" a VM via virt-manager showed only a black screen with the CPU usage graphic in virt-manager at a steady state. No error messages, only a black screen.
If no one else has seen this I will look to the cent-virt mailing list as suggested. I am assuming however, that this is a CR repo issue.
Emmett
kernel-2.6.32-131.17.1.el6.x86_64 (after update and currently)
Maybe you're running into this issue?
http://lists.centos.org/pipermail/centos-virt/2011-November/002713.html
On 11/29/2011 01:47 AM, Lars Hecking wrote:
kernel-2.6.32-131.17.1.el6.x86_64 (after update and currently)
Maybe you're running into this issue?
http://lists.centos.org/pipermail/centos-virt/2011-November/002713.html
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
I doubt that. All VMs are booting now with the latest CR kernel (2.6.32-131.17.1). And it appears that bug is about xen not qemu.
I will clone the host and experiment some to narrow down the problem to a specific package. Right now none of the CR repo libvirt or qemu packages are installed on the host. So I really do not know for sure which is the bad guy.
Emmett
On 11/28/2011 10:38 AM, Emmett Culley wrote:
I finally decided to install the CR repo on one of my CentOS 6 machines that I use as a host for some VMs (also running CentOS 6).
Before updating the host I updated a VM that was not critical to test the process and was able to boot that VM, so I went ahead with updating the host and the two VMs that were in use.
When I rebooted the host none of the VM would start.
Once I figured out that it had to be the host I started downgrading the virtual packages. I was finally able to get the VMs to boot after dwongrading qemu-kvm.
Has anyone else seen this?
Emmett _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Not sure if this is related, but I have a redhat server and after I upgraded from 6.0 to 6.1 libvirtd fails to start, probably due to some dependancy on another process which is not running yet, but when I start it manually after initial boot, it comes up. I have not had time to troubleshoot further.
Nataraj
-
Akemi Yagi -
Emmett Culley -
Lars Hecking -
Nataraj