[lxc-users] Question about Live Migration

[Tycho Andersen]

On Tue, Feb 10, 2015 at 11:51:05AM +0100, Thouraya TH wrote:
> Hello all,
> 
> *1)*
> root at g-71:~# lxc-ls -f
> NAME     STATE    IPV4  IPV6  GROUPS  AUTOSTART
> -----------------------------------------------
> worker   STOPPED  -     -     -       NO
> 
> root at g-71:~# cd /var/lib/lxc/worker
> 
> 
> *root at g-71:/var/lib/lxc/worker# du -sh575M    .*
> Is that the size of the container ?
> 
> *2) http://tycho.ws/blog/2014/09/container-migration.html
> <http://tycho.ws/blog/2014/09/container-migration.html>*
> curl -s $(sudo lxc-info -n worker2 -H -i) didn't work for, i think that is
> because i'm behind a proxy!  Have you an idea please ?
> 
> i have these messages:
> 
> <H1>ERROR</H1>
> <H2>*The requested URL could not be retrieved*</H2>
> <HR noshade size="1px">
> <P>
> *While trying to retrieve the URL:*
> <A HREF="http://10.0.3.30/">http://10.0.3.30/</A>
> <P>
> The following error was encountered:
> <UL>
> <LI>

You might check your network routes, but it is a moot point anyway
because the bit below isn't working.

> 
> *3)* Despite the error, i have done the migration of the container
> On the destination node, i have find the container stopped
> worker2  STOPPED  -           -     -       NO
> 
> that's it ? or i have to find it running ?

It should be running, so something didn't work. There is probably a
log in /tmp/checkpoint/restore.log with some information about why it
failed. To head off another round trip, it probably failed because
criu doesn't support fuse modules, and by default these days the daily
build of lxc has lxcfs included in it, which is a fuse filesystem.

The quickest way to work around this is probably just to uninstall
lxcfs on the host.

Tycho

> Thanks a lot for help.
> Best regards.

> _______________________________________________
> lxc-users mailing list
> lxc-users at lists.linuxcontainers.org
> http://lists.linuxcontainers.org/listinfo/lxc-users