Dear all:
I am trying to migrate a WinXP HVM (64M memory) from a machine
(sender) to another machine (receiver) through AoE disk, however,
there are some errors when do the following command:
(My xen version is xen-4.0.1-rc5.hg, kernel 2.6.31.13, xen xen-4.0.1-rc6-pre,
Both sender and receiver are the same xen version)
xm migrate --live domain_id receiver_ip
============================================== The error message at sender
side is:
Error: timed out
Usage: xm migrate <Domain> <Host>
...
==============================================
Then, I read the xend.log (sender) and find this error message:
==============================================[2010-11-20 18:12:45 4667] INFO
(XendDomain:1206) Domain
WinXP_HVM_Migrate (1) unpaused.
[2010-11-20 18:14:06 4667] DEBUG (XendCheckpoint:124) [xc_save]:
/usr/lib/xen/bin/xc_save 35 1 0 0 5
[2010-11-20 18:14:06 4667] INFO (XendCheckpoint:423) xc_save: failed
to get the suspend evtchn port
...
[2010-11-20 18:14:33 4667] WARNING (image:552) domain
migrating-WinXP_HVM_Migrate: device model failure: pid 4871: died due
to signal 6; see /var/log/xen/qemu-dm-WinXP_HVM_Migrate.log
[2010-11-20 18:14:33 4667] INFO (XendDomainInfo:2088) Domain has
shutdown: name=migrating-WinXP_HVM_Migrate id=1 reason=suspend.
[2010-11-20 18:14:43 4667] INFO (XendCheckpoint:423) ERROR Internal
error: Suspend request failed
[2010-11-20 18:14:43 4667] INFO (XendCheckpoint:423) ERROR Internal
error: Domain appears not to have suspended
[2010-11-20 18:14:48 4667] INFO (XendCheckpoint:423) xc_save: timed
out waiting for qemu logdirty response.
...
==============================================
And the qemu-dm.log gives a "buffer overflow" message as
following:
==============================================...
device model saving state
*** buffer overflow detected ***: /usr/lib64/xen/bin/qemu-dm terminated
======= Backtrace: ========/lib/libc.so.6(__fortify_fail+0x37)[0x7fecab467217]
/lib/libc.so.6(+0xfe0d0)[0x7fecab4660d0]
/lib/libc.so.6(+0xfd539)[0x7fecab465539]
/lib/libc.so.6(_IO_default_xsputn+0xcc)[0x7fecab3ddd1c]
/lib/libc.so.6(_IO_vfprintf+0x14c)[0x7fecab3ad4ec]
==============================================
The logs are also attached. Any ideas or patches to fix this problem ?
Thanks!
Best Regards,
Tang-Hsun Tu
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel
Did you modify the xend-config.sxp correctly? Those comments in the
discussion thread below, could be helpful for your question.
http://lists.xensource.com/archives/html/xen-users/2009-03/msg00079.html
thanks,
-James ( Song Wei)
<< : xen-devel-bounces@lists.xensource.com
[mailto:xen-devel-bounces@lists.xensource.com]
<< xen-devel@lists.xensource.com
<<[Xen-devel] Xen Live Migration Fail
Dear all:
I am trying to migrate a WinXP HVM (64M memory) from a machine
(sender) to another machine (receiver) through AoE disk, however, there are
some errors when do the following command:
(My xen version is xen-4.0.1-rc5.hg, kernel 2.6.31.13, xen
xen-4.0.1-rc6-pre, Both sender and receiver are the same xen version)
xm migrate --live domain_id receiver_ip
============================================== The error message at sender
side is:
Error: timed out
Usage: xm migrate <Domain> <Host>
...
==============================================
Then, I read the xend.log (sender) and find this error message:
==============================================[2010-11-20 18:12:45 4667] INFO
(XendDomain:1206) Domain WinXP_HVM_Migrate
(1) unpaused.
[2010-11-20 18:14:06 4667] DEBUG (XendCheckpoint:124) [xc_save]:
/usr/lib/xen/bin/xc_save 35 1 0 0 5
[2010-11-20 18:14:06 4667] INFO (XendCheckpoint:423) xc_save: failed to get
the suspend evtchn port ...
[2010-11-20 18:14:33 4667] WARNING (image:552) domain
migrating-WinXP_HVM_Migrate: device model failure: pid 4871: died due to
signal 6; see /var/log/xen/qemu-dm-WinXP_HVM_Migrate.log
[2010-11-20 18:14:33 4667] INFO (XendDomainInfo:2088) Domain has
shutdown: name=migrating-WinXP_HVM_Migrate id=1 reason=suspend.
[2010-11-20 18:14:43 4667] INFO (XendCheckpoint:423) ERROR Internal
error: Suspend request failed
[2010-11-20 18:14:43 4667] INFO (XendCheckpoint:423) ERROR Internal
error: Domain appears not to have suspended [2010-11-20 18:14:48 4667] INFO
(XendCheckpoint:423) xc_save: timed out waiting for qemu logdirty response.
...
==============================================
And the qemu-dm.log gives a "buffer overflow" message as
following:
==============================================...
device model saving state
*** buffer overflow detected ***: /usr/lib64/xen/bin/qemu-dm terminated
======= Backtrace: ========/lib/libc.so.6(__fortify_fail+0x37)[0x7fecab467217]
/lib/libc.so.6(+0xfe0d0)[0x7fecab4660d0]
/lib/libc.so.6(+0xfd539)[0x7fecab465539]
/lib/libc.so.6(_IO_default_xsputn+0xcc)[0x7fecab3ddd1c]
/lib/libc.so.6(_IO_vfprintf+0x14c)[0x7fecab3ad4ec]
==============================================
The logs are also attached. Any ideas or patches to fix this problem ?
Thanks!
Best Regards,
Tang-Hsun Tu
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel
Hi,
Thanks for your reply.
In recent days, I update my xen to the latest xen-4.0-testing.hg,
and retry the migration. It works fine.
However, it still has some strange problems.
Supporse there are two machines, A and B.
If I start a DomU (HVM) at A, then migrate to B, it works.
Then, I migrate from B to A, it works.
However, when I start a DomU at B, then migrate to A, it fails.
It means, if the first migration from B to A, it fails.
But, A->B, then B->A, it works.
It fails at restoring the DomU, and
the log shows "ERROR Internal error: error setting the HVM context".
The logs are also attached.
==================================================================
Another problem is that when I patch QEMU to support multiple DomUs to
play audio (enable mix),
like this,
the patch at tools/xen-setup,
#./configure --disable-gfx-check --disable-curses --disable-slirp "$@"
--prefix=${PREFIX}
./configure --audio-drv-list=alsa --enable-mixemu --disable-gfx-check
--disable-curses --disable-slirp "$@" --prefix=${PREFIX}
Then, the latest xen also fails, does this patch will affect live migration?
Thanks!
Best Regards,
Tang-Hsun Tu
於 2010/11/23 上午 10:05, songwei 提到:> Did you modify the xend-config.sxp correctly? Those comments in the
> discussion thread below, could be helpful for your question.
> http://lists.xensource.com/archives/html/xen-users/2009-03/msg00079.html
>
> thanks,
> -James ( Song Wei)
>
>
>
> sender: xen-devel-bounces@lists.xensource.com
> [mailto:xen-devel-bounces@lists.xensource.com] represent 涂小訓
> timer: 2010年11月20日 18:49
> receiver: xen-devel@lists.xensource.com
> subject: [Xen-devel] Xen Live Migration Fail
>
> Dear all:
>
> I am trying to migrate a WinXP HVM (64M memory) from a machine
> (sender) to another machine (receiver) through AoE disk, however, there are
> some errors when do the following command:
> (My xen version is xen-4.0.1-rc5.hg, kernel 2.6.31.13, xen
> xen-4.0.1-rc6-pre, Both sender and receiver are the same xen version)
>
> xm migrate --live domain_id receiver_ip
>
> ==============================================> The error message
at sender side is:
>
> Error: timed out
> Usage: xm migrate <Domain> <Host>
> ...
> ==============================================>
>
> Then, I read the xend.log (sender) and find this error message:
>
> ==============================================> [2010-11-20 18:12:45
4667] INFO (XendDomain:1206) Domain WinXP_HVM_Migrate
> (1) unpaused.
> [2010-11-20 18:14:06 4667] DEBUG (XendCheckpoint:124) [xc_save]:
> /usr/lib/xen/bin/xc_save 35 1 0 0 5
> [2010-11-20 18:14:06 4667] INFO (XendCheckpoint:423) xc_save: failed to get
> the suspend evtchn port ...
> [2010-11-20 18:14:33 4667] WARNING (image:552) domain
> migrating-WinXP_HVM_Migrate: device model failure: pid 4871: died due to
> signal 6; see /var/log/xen/qemu-dm-WinXP_HVM_Migrate.log
> [2010-11-20 18:14:33 4667] INFO (XendDomainInfo:2088) Domain has
> shutdown: name=migrating-WinXP_HVM_Migrate id=1 reason=suspend.
> [2010-11-20 18:14:43 4667] INFO (XendCheckpoint:423) ERROR Internal
> error: Suspend request failed
> [2010-11-20 18:14:43 4667] INFO (XendCheckpoint:423) ERROR Internal
> error: Domain appears not to have suspended [2010-11-20 18:14:48 4667] INFO
> (XendCheckpoint:423) xc_save: timed out waiting for qemu logdirty response.
> ...
> ==============================================>
>
> And the qemu-dm.log gives a "buffer overflow" message as
following:
>
> ==============================================> ...
> device model saving state
> *** buffer overflow detected ***: /usr/lib64/xen/bin/qemu-dm terminated
> ======= Backtrace: ========>
/lib/libc.so.6(__fortify_fail+0x37)[0x7fecab467217]
> /lib/libc.so.6(+0xfe0d0)[0x7fecab4660d0]
> /lib/libc.so.6(+0xfd539)[0x7fecab465539]
> /lib/libc.so.6(_IO_default_xsputn+0xcc)[0x7fecab3ddd1c]
> /lib/libc.so.6(_IO_vfprintf+0x14c)[0x7fecab3ad4ec]
> ==============================================>
>
> The logs are also attached. Any ideas or patches to fix this problem ?
> Thanks!
>
> Best Regards,
> Tang-Hsun Tu
>
>
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel
Seemingly Similar Threads
- Internal error during live migration saving
- Live migration problem with xen 3.2
- soft lockups during live migrate..
- Snapshot fail, when snapshot a vm the second time. (already update to xen-4.0.1 and kernel-2.6.32.25)
- Snapshot fail, when snapshot a vm the second time. (already update to xen-4.0.1 and kernel-2.6.32.25)