virt-v2v: "Didn't find a grub entry for kernel version ..."
by Ask Bjørn Hansen
Hi everyone,
I'm trying to convert a CentOS (5.5) xen virtual machine to a KVM based one. I exported the XML, copied it and the device data (LVM partitions on both boxes); adjusted the disk device path and the name of the bridge in the XML and then ran:
virt-v2v -f /etc/virt-v2v.conf -i libvirtxml mongo1.xml
My virt-v2v.conf is below.
The command exits with "virt-v2v: Didn't find a grub entry for kernel version 2.6.18-194.8.1.el5" -- how do I fix that?
The virt-v2v version is 0.4.0-1.
- ask
<virt-v2v>
<path-root>/var/lib/virt-v2v/software</path-root>
<iso-path>/var/lib/virt-v2v/transfer.iso</iso-path>
<app os='centos' name='kernel' major='5' minor='5' arch='x86_64'>
<path>centos/5/kernel-2.6.18-194.8.1.el5.x86_64.rpm</path>
<dep>centos/5/ecryptfs-utils-75-5.el5.x86_64.rpm</dep>
<dep>centos/5/lvm2-2.02.56-8.el5_5.6.x86_64.rpm</dep>
<dep>centos/5/device-mapper-1.02.39-1.el5_5.2.x86_64.rpm</dep>
<dep>centos/5/device-mapper-event-1.02.39-1.el5_5.2.x86_64.rpm</dep>
</app>
<!-- Networks -->
<!-- The default Xen bridge name -->
<network type='bridge' name='xenbr1'>
<network type='network' name='default'/>
</network>
<!-- The default ESX bridge name -->
<network type='bridge' name='VM Network'>
<network type='network' name='default'/>
</network>
</virt-v2v>
13 years, 8 months
Slow virtio network startup?
by Tom Horsley
I just replaced my block and network virtio drivers in my
Windows XP machine with the newest ones (the old block
driver would bluescreen at boot time). The new drivers
seem to work, but the network takes almost a minute to
get initialized. During this time if I try to do things
like right click on network and bring up properties,
nothing happens at all, the window with the list of
interfaces doesn't even appear for about a minute, and
when it does appear, it is blank - I have to right click
and refresh before the network shows up. Once this finally
happens, it seems to be working normally. I can tell
it to repair, and it only takes a few seconds to renew
the DHCP lease, so I don't think this is a DHCP problem.
When I was using the older virtio drivers on the older
libvirt, the network would come up right away without
this mysterious delay, and other than updating the
virtio drivers I haven't changed any settings.
Anyone have any clues to the problem, or ideas for
how I can tell what is going on?
13 years, 8 months
[OT?] Help debugging qemu-f13-host/centos-guest issue when SMP is enabled.
by Gilboa Davara
Hello all,
I know that this question is beyond the scope of this ML, but maybe
someone can help me.
Recently I've started experiencing client hangs across the board on SMP
CentOS 5.5/x86_64 clients running under a number of F13/x86_64 hosts
when -smp is enabled [1].
Looking at the yum logs on both the hosts and the guest doesn't seem to
point at any recent update that might have broken anything.
As we use CentOS guests to test our application before pushing updates
down to our RHEL servers, the lack of SMP support is very problematic.
I did my best to generate an informative bug report, but I fear that
lacking additional information, this bug report has little chance of
getting solved within the F13 timeframe.
So, how can determine if the problem is host based or guest based? (The
CentOS 5.5/x86_64 guests were rock solid until ~2 weeks ago with smp=2
and up to smp=8)
How can help the qemu maintainers triage this bug?
- Gilboa
[1] https://bugzilla.redhat.com/show_bug.cgi?id=619306
13 years, 8 months
Re: [fedora-virt] Suddenly KVM/virt-manager is failing on qcow2 images
by Scott Dowdle
Greetings,
Opps, I replied back to the original within 5 minutes of it appearing on the list but I guess my reply did NOT include the fedora-virt list and only went directly to the original poster.
Anyhoo... we already have a fix for it and it is listed below.
- - - - -
----- Original Message -----
> Hi, today all my KVMs based on qcow2 images have failed with "not a
> bootable hard disk" message.
>
> I am running F13 x64 fully yum updated on AMD Phenom II X4.
>
> All these VMs were working fine (linux,freebsd,winxp).
>
> When I booted a VM from an ISO image and looked at the /dev/vda -
> there was a "corrupt" partition table.
>
> When I converted the qcow2 image to raw, and looked at the partition
> table, it was fine.
>
> When I switched the VM to use the raw image, it booted and ran fine.
>
> Any known issues with recent KVM/qcow2 interfaction?
>
> Packages:
> kernel-2.6.33.6-147.fc13.x86_64
> gpxe-roms-qemu-1.0.0-1.fc13.noarch
> qemu-common-0.12.3-8.fc13.x86_64
> qemu-img-0.12.3-8.fc13.x86_64
> qemu-system-x86-0.12.3-8.fc13.x86_64
> qemu-kvm-0.12.3-8.fc13.x86_64
>
> Help!? Tks.
Yeah, I had the same problem and thanks goes to nirik for pointing me to the following info that fixed it for me:
https://admin.fedoraproject.org/updates/libvirt-0.8.2-1.fc13
Read the notes. What I did was edit the config and turn switch probing back... and edit my VM's config and change the disk type from "raw" to "qcow2".
- - - - -
TYL,
--
Scott Dowdle
704 Church Street
Belgrade, MT 59714
(406)388-0827 [home]
(406)994-3931 [work]
13 years, 9 months
Suddenly KVM/virt-manager is failing on qcow2 images
by Richard Chan
Hi, today all my KVMs based on qcow2 images have failed with "not a bootable
hard disk" message.
I am running F13 x64 fully yum updated on AMD Phenom II X4.
All these VMs were working fine (linux,freebsd,winxp).
When I booted a VM from an ISO image and looked at the /dev/vda - there was
a "corrupt" partition table.
When I converted the qcow2 image to raw, and looked at the partition table,
it was fine.
When I switched the VM to use the raw image, it booted and ran fine.
Any known issues with recent KVM/qcow2 interfaction?
Packages:
kernel-2.6.33.6-147.fc13.x86_64
gpxe-roms-qemu-1.0.0-1.fc13.noarch
qemu-common-0.12.3-8.fc13.x86_64
qemu-img-0.12.3-8.fc13.x86_64
qemu-system-x86-0.12.3-8.fc13.x86_64
qemu-kvm-0.12.3-8.fc13.x86_64
Help!? Tks.
13 years, 9 months
Hey, my Win XP machine now bluescreens
by Tom Horsley
I've had a Windows XP KVM running through several fedora
upgrades and updates, and I just tried starting it for
the first time since getting the libvirt-0.8.2-1.fc13.x86_64
update, and now it bluescreens.
I had been running the redhat virtio disk and network drivers
inside the KVM, have they gone incompatible on me? (I guess
I could try switching the KVM back to IDE emulation and
see if it starts working).
13 years, 9 months
Re: [fedora-virt] Fedora Virt Status
by Scott Dowdle
Greetings,
----- Original Message -----
> > I'm wondering if there will be a test repo for Fedora 13 because it
> > would be nice to get SPICE going in Fedora 13 too!
>
> Actually there is one, which works quite well for me with F13 :
>
> $ cat /etc/yum.repos.d/spice-unstable.repo
> [spice-unstable]
> name=spice development bits for fedora $releasever
> baseurl=http://kraxel.fedorapeople.org/spice/fedora$releasever
> enabled=1
> gpgcheck=0
>
> In particular, the "qemu-spice" package provides a new
> "/usr/bin/qemu-spice" binary (since it's not fully merged with
> upstream qemu, both qemu and qemu-spice can be installed side by side
> this way).
Hmm, thanks for the info. I gave it a try but so far I'm stumped. I've not really used qemu/kvm from the command line much and mostly use virt-manager. I read the instructions here:
http://kraxel.fedorapeople.org/spice/README
I got the following packages installed (from my /var/log/yum.log)
Jul 16 11:37:25 Installed: celt051-0.5.1.3-2.fc13.x86_64
Jul 16 11:37:26 Installed: spice-server-0.5.2-2.fc13.x86_64
Jul 16 11:37:29 Installed: qemu-spice-0.12.50-0.git.20100708.fc13.x86_64
Jul 16 11:37:30 Installed: spice-client-0.5.2-2.fc13.x86_64
Then I ran a VM from virt-manager and did a "ps auxwww | grep kvm" to see what the full command virt-manager uses. I took that command and modified it for qemu-spice and here is what I came up with:
/usr/bin/qemu-spice \
-spice port=1234,password=mypassword \
-device qxl \
-vga qxl \
-S \
-M pc \
-cpu core2duo,+x2apic \
-enable-kvm \
-m 2048 \
-smp 2,sockets=2,cores=1,threads=1 \
-name montanalinux32 \
-uuid 9cf9479c-3476-9aa5-bdd3-dccf5d8af015 \
-nodefaults \
-chardev socket,id=monitor,path=/var/lib/libvirt/qemu/montanalinux32.monitor,server,nowait \
-mon chardev=monitor,mode=readline \
-rtc base=utc \
-boot c \
-drive file=/vm/montanalinux32.img,if=none,id=drive-virtio-disk0,boot=on,format=raw \
-device virtio-blk-pci,bus=pci.0,addr=0x4,drive=drive-virtio-disk0,id=virtio-disk0 \
-drive if=none,media=cdrom,id=drive-ide0-1-0 \
-device ide-drive,bus=ide.1,unit=0,drive=drive-ide0-1-0,id=ide0-1-0 \
-device virtio-net-pci,vlan=0,id=net0,mac=52:54:00:3a:0d:08,bus=pci.0,addr=0x5 \
-net tap,fd=49,vlan=0,name=hostnet0 \
-chardev pty,id=serial0 \
-device isa-serial,chardev=serial0 \
-usb \
-device usb-tablet,id=input0 \
-k en-us \
-device AC97,id=sound0,bus=pci.0,addr=0x6
That seems to start up (at least it shows in ps output) BUT when I try to connect to it with the spice client via...
spicec -h localhost -p 1234 -w mypassword
...all I get is a black display in the window and it doesn't connect.
Please note that I used an existing Fedora 13 KVM VM. If I understand correctly the VM needs to have the qxl driver installed in it so maybe that is why it isn't working. Or do the command line options given above inject the qxl driver into the VM magically?
I'm very new to this so I have little idea what I'm doing much less what I'm doing wrong. Any help would be appreciated.
Thanks,
--
Scott Dowdle
704 Church Street
Belgrade, MT 59714
(406)388-0827 [home]
(406)994-3931 [work]
13 years, 9 months
Re: [fedora-virt] virtio-win package?
by Andrew Cathrow
----- "Tom Horsley" <horsley1953(a)gmail.com> wrote:
> From: "Tom Horsley" <horsley1953(a)gmail.com>
> To: virt(a)lists.fedoraproject.org
> Sent: Sunday, August 1, 2010 5:33:52 PM GMT -05:00 US/Canada Eastern
> Subject: Re: [fedora-virt] virtio-win package?
>
> On Sun, 01 Aug 2010 15:53:22 -0500
> Justin M. Forbes wrote:
>
> > The documentation is busted. We are not allowed to include virtio-win
> > drivers in Fedora repositories until we can build them without Windows.
>
> I figured as much, but just thought I'd ask. The new drivers
> do seem to work, but there doesn't appear to be any direct
> way to install them so I can boot from virtio without taking
> a round-about approach.
Most direct approach is to install the VM with both the XP CD rom and the virtual floppy disk from here http://alt.fedoraproject.org/pub/alt/virtio-win/13/images/bin/viostor-31-...
Set the disk to be virtio and hit the key to load driver disk during install IIRC it's F6
>
> What finally worked was:
>
> Make boot disk IDE
>
> Make a 2nd disk and call it virtio
>
> Boot XP and let the new hardware wizard ask about the
> new disk.
>
> Point it to the CDROM with the guest drivers and let
> it install the virtio disk driver.
>
> Now I can shutdown the XP KVM, fiddle the xml file
> to change the boot disk back to virtio and remove
> the 2nd dummy disk image.
>
> Finally I can boot from virtio (whew! :-).
> _______________________________________________
> virt mailing list
> virt(a)lists.fedoraproject.org
> https://admin.fedoraproject.org/mailman/listinfo/virt
>
13 years, 9 months
