[Xen-users] Live migration problem

[Alain BARTHE <alain.barthe@xxxxxxxx>]

Hi,

We have a dozen of dom0, running about
50 domUs. All domUs have disks on LVM, disks are located on a SAN, so it
is very easy to live migrate simply by attaching a domU disk on two dom0s.


All dom0s run CentOS50 with Xen 3.0.3.

We tried many time to migrate in vitro
some little domUs and it worked perfectly.

Now we really need to migrate a big
domU (big by its RAM: 3.6 GB) and we have the following error (log on the
from dom0):

[2008-01-10 10:38:08 xend 30141] DEBUG
(balloon:127) Balloon: 18520 KiB free; need 17408; done.
[2008-01-10 10:38:09 xend 30141] DEBUG
(XendCheckpoint:83) [xc_save]: /usr/lib/xen/bin/xc_save 23 1 0 0 1
[2008-01-10 10:38:09 xend 30141] INFO
(XendCheckpoint:239) Couldn't enable shadow mode: 12
[2008-01-10 10:38:09 xend 30141] INFO
(XendCheckpoint:239) Save exit rc=1
[2008-01-10 10:38:09 xend 30141] ERROR
(XendCheckpoint:106) Save failed on domain ctc31pinl (1).
Traceback (most recent call last):
  File "/usr/lib/python2.4/site-packages/xen/xend/XendCheckpoint.py",
line 100, in save
    forkHelper(cmd, fd, saveInputHandler,
False)
  File "/usr/lib/python2.4/site-packages/xen/xend/XendCheckpoint.py",
line 227, in forkHelper
    raise XendError("%s
failed" % string.join(cmd))
XendError: /usr/lib/xen/bin/xc_save
23 1 0 0 1 failed

Google said that this error may occur
when the 2 dom0s have different CPUs. It was the case at our first trial,
we took another dom0 with a rigourously identical hardware and the error
remains the same.
We checked that the target dom0 has
enough RAM, and that /var partition is big enough on both dom0s.

Thanks for help.
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users