Proxmox VE 1.8 released!

Well, when sda2, on which is the /boot/ partition is now sdc2, that looks kind of aproblem to me when the os quits its ramdisk.

And with 3 NICs, two of which are 1Gb/s able and one only 100Mb/s, when eth0 becomes eth2, all the virtual machine lack a bit of bandwidth. If you see what I mean.

thats why we use UUID pointing to the boot partition instead of /sdx ...

Code:
cat /etc/fstab

# <file system> <mount point> <type> <options> <dump> <pass>
/dev/pve/root / ext3 errors=remount-ro 0 1
/dev/pve/data /var/lib/vz ext3 defaults 0 1
UUID=c256ea93-bc9a-42de-83cc-56f10c1f2d6d /boot ext3 defaults 0 1
/dev/pve/swap none swap sw 0 0
proc /proc proc defaults 0 0
 
... just in the case you did not know it - we do not recommend or support the use of soft raid.
 
Hi,

after updating to the latest version (incl. a reboot of the host), my KVM running Ubuntu does not boot anymore - directly after trying to boot from the harddisk, the following occurs on VNC screen.

unaligned pointer 0x1020101 (number is changing)
Aborted. Press any key to exit.

Pressing a key ends up that the first line is deleted and the 2nd either appears again or it does not react anymore.

pve-manager: 1.8-15 (pve-manager/1.8/5754)
running kernel: 2.6.32-4-pve
proxmox-ve-2.6.32: 1.8-32
pve-kernel-2.6.32-4-pve: 2.6.32-32
qemu-server: 1.1-30
pve-firmware: 1.0-11
libpve-storage-perl: 1.0-16
vncterm: 0.9-2
vzctl: 3.0.24-1pve4
vzdump: 1.2-11
vzprocps: 2.0.11-1dso2
vzquota: 3.0.11-1
pve-qemu-kvm: 0.14.0-2
ksm-control-daemon: 1.0-5

Any ideas?
 
thats why we use UUID pointing to the boot partition instead of /sdx ...

Yes, indeed. .. Obviously. Shame on me not having been used to that.:(

Do you have a similar solution for the NICs ?
 
try downgrading the pve-qemu-kvm package and try again.

download the 1.7 package:

Code:
wget ftp://download.proxmox.com/debian/dists/lenny/pve1.7/binary-amd64/pve-qemu-kvm_0.13.0-3_amd64.deb
and install:
Code:
dpkg -i pve-qemu-kvm_0.13.0-3_amd64.deb
no reboot of the host needed, just stop/start of all KVM guests.

do you run the latest kernel inside your KVM guest?
 
Well, when sda2, on which is the /boot/ partition is now sdc2, that looks kind of aproblem to me when the os quits its ramdisk.

please use UUID in fstab

And with 3 NICs, two of which are 1Gb/s able and one only 100Mb/s, when eth0 becomes eth2, all the virtual machine lack a bit of bandwidth. If you see what I mean.

'udev' prevents that, so this can't happen (devices are renamed by udev).
 
Thanks Tom, that did it for me until i'd update again ;-)

The kernel inside Ubuntu 10.10 should be up to date - at least I dont get any update offers (2.6.35-28-generic)

Anything you want me to test?
 
post your VMID.conf file from the KVM guest, lets see if I can reproduce it.
 
name: desktop
ide2: none,media=cdrom
bootdisk: ide0
ostype: l26
ide0: local:202/vm-202-disk-1.raw
memory: 1024
sockets: 1
vlanu: rtl8139=CE:BD:BC:25:13:40
onboot: 0
cores: 1
args: -vga vmware
 
I am pretty sure its vmware vga related, remove this line and test again. can you confirm this?
 
Yes, confirmed - its working now with the latest updates installed.

But please note that this line was not added by hand - seems that Proxmox GUI added it at time of KVM creation.
 
Yes, confirmed - its working now with the latest updates installed.

thanks for testing.

But please note that this line was not added by hand - seems that Proxmox GUI added it at time of KVM creation.

no, thats just not possible. the gui never adds 'args:'
 
No problem - its just a test KVM so it has fulfilled its purpose :-)

Ok, maybe I added it to fix another issue but I dont remember anymore (maybe I should add a SVN to keep track of my changes ;) )
 
Do you have a similar solution for the NICs ?

I may be wildly off track, but recently ran into what seems to be the same problem on my Proxmox VE 1.7 system that you are describing here, albeit in a completely different context. Essentially my four ethernet NIC's were renumbered from eth0, eth1, eth2, eth3 to eth4, eth5, eth6, eth7, thereby causing numerous problems.

The solution was to delete:

/etc/udev/rules.d/70-persistent-net.rules

For your sanity, you may want to back up this file first, though I don't believe deleting it will cause any damage. YMMV, good luck!

- Acorp
www.acorp.net
 
Does Proxmox VE 1.8 support Kernel Samepage Merging (KSM) out of the box? I just installed it, but when I look, I don't see it enabled:

proxmox:/sys/kernel/mm# grep KSM /boot/config-2.6.32-4-pve
# CONFIG_KSM is not set
proxmox:/sys/kernel/mm# cd /sys/kernel/mm/
proxmox:/sys/kernel/mm# ls
hugepages
(no ksm subdir)

so do I need to follow the instructions at http://pve.proxmox.com/wiki/Proxmox_VE_Kernel and run:

aptitude update
aptitude safe-upgrade
aptitude install proxmox-ve-2.6.35

to install the 2.6.35 kernel with KSM support?
 
I couldn't get kernel 2.6.35 to install, but I reran
aptitude update
aptitude safe-upgrade
aptitude install proxmox-ve-2.6.35

again and then it installed 2.6.35 fine. thanks
 
Last edited:
Hi to all

Proxmox VEe is a great product and never gave me problems from more than a year.

Unfortunately in the last case I had the same problemf of gbrucker with a guest ubuntu server (Zential) and I solved the problem reading your help the to him.

After this, I updated the guest to the last kernerl

Linux server 2.6.32-30-generic-pae #59-Ubuntu SMP Tue Mar 1 23:01:33 UTC 2011 i686 GNU/Linux

and after I upgraded again proxmox.

The guest don't want to start and give back with the same error. The only way was to roll back to the previous configuration pve-qemu-kvm_0.13.0-3.


This is my VMID,conf

name: atlante
ide2: none,media=cdrom
bootdisk: ide0
ostype: l26
ide0: local:105/vm-105-disk-1.raw
memory: 2048
onboot: 1
sockets: 2
boot: cad
freeze: 0
cpuunits: 1000
vga: std
acpi: 1
kvm: 1
cores: 2
vlan0: rtl8139=9E:3C:CC:EE:AA:8B
vlan1: rtl8139=EE:19:D4:D5:FC:4F


Piero Proietti
 
Last edited: