[pve-devel] [Qemu-devel] QEMU LIve Migration - swap_free: Bad swap file entry

Stefan Priebe - Profihost AG s.priebe at profihost.ag
Thu Feb 6 11:49:26 CET 2014


Am 06.02.2014 11:22, schrieb Orit Wasserman:
> On 02/06/2014 09:20 AM, Stefan Priebe - Profihost AG wrote:
>> Am 05.02.2014 21:15, schrieb Dr. David Alan Gilbert:
>>> * Stefan Priebe (s.priebe at profihost.ag) wrote:
>>>> Hello,
>>>>
>>>> after live migrating machines with a lot of memory (32GB, 48GB, ...)
>>>> i see pretty often crashing services after migration and the guest
>>>> kernel prints:
>>>>
>>>> [1707620.031806] swap_free: Bad swap file entry 00377410
>>>> [1707620.031806] swap_free: Bad swap file entry 00593c48
>>>> [1707620.031807] swap_free: Bad swap file entry 03201430
>>>> [1707620.031807] swap_free: Bad swap file entry 01bc5900
>>>> [1707620.031807] swap_free: Bad swap file entry 0173ce40
>>>> [1707620.031808] swap_free: Bad swap file entry 011c0270
>>>> [1707620.031808] swap_free: Bad swap file entry 03c58ae8
>>>> [1707660.749059] BUG: Bad rss-counter state mm:ffff88064d09f380
>>>> idx:1 val:1536
>>>> [1707660.749937] BUG: Bad rss-counter state mm:ffff88064d09f380
>>>> idx:2 val:-1536
>>>>
> 
> Is this live migration with shared storage? what kind of shared storage?

Yes - i'm using ceph / rbd.

> Does this happens with smaller guests?

Never seen that. Always with guest having memory > 16GB.

>>>> Qemu is 1.7
>>>>
>>>> Does anybody know a fix?
>>>
>>> I don't, but some more information about:
>>>      1) What guest you're running
>>
>> Linux guest the output is also from the guest. Kernel 3.10.26
>>
>>>      2) The configuration of your hosts
>>
>> What do you mean by that?
>>
>>>      3) The command line (or XML if you're running libvirt) for
>>>         your qemu so we can see what devices you're running.
>>
>> qemu -chardev
>> socket,id=qmp,path=/var/run/qemu-server/179.qmp,server,nowait -mon
>> chardev=qmp,mode=control -vnc
>> unix:/var/run/qemu-server/179.vnc,x509,password -pidfile
>> /var/run/qemu-server/179.pid -daemonize -name K31953 -smp
>> sockets=1,cores=16 -nodefaults -boot
>> menu=on,strict=on,reboot-timeout=1000 -vga cirrus -cpu
>> kvm64,+lahf_lm,+x2apic,+sep -k de -m 32768 -device
>> piix3-usb-uhci,id=uhci,bus=pci.0,addr=0x1.0x2 -device
>> usb-tablet,id=tablet,bus=uhci.0,port=1 -device
>> virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 -drive
>> if=none,id=drive-ide2,media=cdrom,aio=native -device
>> ide-cd,bus=ide.1,unit=0,drive=drive-ide2,id=ide2,bootindex=200 -device
>> virtio-scsi-pci,id=scsihw0,bus=pci.0,addr=0x5 -drive
>> file=rbd:...,if=none,id=drive-scsi0,iops_rd=1000,iops_wr=500,bps_rd=314572800,bps_wr=209715200,aio=native,discard=on
>>
>> -device
>> scsi-hd,bus=scsihw0.0,channel=0,scsi-id=0,lun=0,drive=drive-scsi0,id=scsi0,bootindex=100
>>
>> -netdev
>> type=tap,id=net0,ifname=tap179i0,script=/var/lib/qemu-server/pve-bridge,vhost=on
>>
>> -device
>> virtio-net-pci,mac=CA:CA:23:AC:2D:C5,netdev=net0,bus=pci.0,addr=0x12,id=net0,bootindex=300
>>
>> -rtc base=localtime -machine type=pc-i440fx-1.7
>>
>>> Do you get any messages on either the source or destination
>>> qemu during the migrate?

no

Stefan



More information about the pve-devel mailing list