Clock goes crazy after migrate

[whinpo]

[SOLVED] Clock goes crazy after migrate

Hi,

I'm using 2 Proxmox-VE servers in cluster.
The first is a Dell Poweredge 2950, the other one a Dell Poweredge 1950

If I start a debian VM on the 2950 and then migrate the VM to the 1950, the clock goes faster (20/30% faster) until I reboot the VM.
Now, if I migrate the new started VM back to the 2950, the clock goes slower...until I reboot...

I would not like adding a crontab with ntpdate every 1 minute...or reboot each time I do a migration

My debian VM is a standard lenny with nothing installed :

debian:~# uname -a
Linux debian 2.6.26-2-686 #1 SMP Sat Dec 26 09:01:51 UTC 2009 i686 GNU/Linux
debian:~#

I've checked as said in http://forum.proxmox.com/threads/2381-Recommended-clock-source-for-KVM-guests

debian:~# cat /sys/devices/system/clocksource/clocksource0/current_clocksource
kvm-clock

My version :
Dell 1950 :

neptune1:~# pveversion -v
pve-manager: 1.5-8 (pve-manager/1.5/4674)
running kernel: 2.6.18-2-pve
proxmox-ve-2.6.18: 1.5-5
pve-kernel-2.6.18-2-pve: 2.6.18-5
qemu-server: 1.1-11
pve-firmware: 1.0-3
libpve-storage-perl: 1.0-10
vncterm: 0.9-2
vzctl: 3.0.23-1pve8
vzdump: 1.2-5
vzprocps: 2.0.11-1dso2
vzquota: 3.0.11-1
pve-qemu-kvm-2.6.18: 0.9.1-5

Dell 2950 :

neptune2:~# pveversion -v
pve-manager: 1.5-8 (pve-manager/1.5/4674)
running kernel: 2.6.18-2-pve
proxmox-ve-2.6.18: 1.5-5
pve-kernel-2.6.18-2-pve: 2.6.18-5
pve-kernel-2.6.24-8-pve: 2.6.24-16
qemu-server: 1.1-11
pve-firmware: 1.0-3
libpve-storage-perl: 1.0-10
vncterm: 0.9-2
vzctl: 3.0.23-1pve8
vzdump: 1.2-5
vzprocps: 2.0.11-1dso2
vzquota: 3.0.11-1
pve-qemu-kvm-2.6.18: 0.9.1-5

the VM conf :

neptune2:~# cat /etc/qemu-server/110.conf 
name: test32bits
ide2: NFSMirror:iso/lenny-debian-504-i386-businesscard.iso,media=cdrom
bootdisk: ide0
ostype: other
ide0: ProxmoxTest-SR:vm-110-disk-1
memory: 512
sockets: 1
vlan0: rtl8139=DE:B5:08:FB:7F:4A

any help would be appreciated

 

[whinpo]

nobody?

 

[valshare]

Hi Whinpo,

i have the same problem with Dell PE850 under Kernel 2.6.18 and Windows XP. The Clock under XP slows down 30%. Only with Kernel 2.6.24 it runs fine here, but with this kernel i can´t do a live migrate

Sorry, that i didn´t can help you.

 

[whinpo]

thanks
that's a first step...
let's try under 2.6.24

 

[whinpo]

tested with 2.6.24.11 => live migration didn't work
tested with 2.6.24.10 => kernel error in VM

[  256.102205] Clocksource tsc unstable (delta = -71848024324 ns)
[  329.120065] BUG: soft lockup - CPU#0 stuck for 68s! [events/0:6]
[  329.120065] Modules linked in: ipv6 loop parport_pc parport snd_pcsp snd_pcm snd_timer snd soundcore psmouse snd_page_alloc serio_raw i2c_piix4 i2c_core button joydev evdev ext3 jbd mbcache ide_cd_mod cdrom usbhid hid ff_memless ide_disk ide_pci_generic 8139too floppy piix 8139cp mii uhci_hcd usbcore ide_core ata_generic libata scsi_mod dock thermal processor fan thermal_sys [last unloaded: scsi_wait_scan]
[  329.120065] 
[  329.120065] Pid: 6, comm: events/0 Not tainted (2.6.26-2-686 #1)
[  329.120065] EIP: 0060:[<c011d60a>] EFLAGS: 00000202 CPU: 0
[  329.120065] EIP is at finish_task_switch+0x26/0x99
[  329.120065] EAX: c1409fc0 EBX: c03bafc0 ECX: c1409fc0 EDX: df9764a0
[  329.120065] ESI: 00000000 EDI: df9764a0 EBP: 00000001 ESP: df44bf5c
[  329.120065]  DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068
[  329.120065] CR0: 8005003b CR2: 09daa8b0 CR3: 1f67f000 CR4: 000006d0
[  329.120065] DR0: 00000000 DR1: 00000000 DR2: 00000000 DR3: 00000000
[  329.120065] DR6: ffff0ff0 DR7: 00000400
[  329.120065]  [<c02b89ce>] ? schedule+0x60c/0x66f
[  329.120065]  [<c012f719>] ? worker_thread+0x0/0xbd
[  329.120065]  [<c012f79f>] ? worker_thread+0x86/0xbd
[  329.120065]  [<c01319d4>] ? autoremove_wake_function+0x0/0x2d
[  329.120065]  [<c0131913>] ? kthread+0x38/0x5d
[  329.120065]  [<c01318db>] ? kthread+0x0/0x5d
[  329.120065]  [<c01044f7>] ? kernel_thread_helper+0x7/0x10
[  329.120065]  =======================

tested under 2.6.32-2 -> seems fixed !