[lxc-users] Kernel lockups when running lxc-start (J?kel)

[Dao Quang Minh]

Yes, i setup the mountpoints in advance. I actually didnt know about the
rootfs trick ( TIL ) :D


On Wed, Mar 12, 2014 at 10:49 PM, Serge Hallyn <serge.hallyn at ubuntu.com>wrote:

> Ok, I see.  I thought you were using something like
>
>         lxc.rootfs =
> aufs:/var/lib/lxc/container-name/rootfs:/var/lib/lxc/container-name/delta
>
> in the container configuration to have lxc do the aufs mount, rather
> than setting the aufs mounts up in advance.
>
> Quoting Daniel, Dao Quang Minh (dqminh89 at gmail.com):
> > Hmm, those entries appear on the host’s /proc/self/mountinfo. I guess
> it’s because we union-mounts on /disk1/container-name and bind mount that
> to /var/lib/lxc/container-name ?
> >
> > Here are the entries for `/proc/self/mountinfo` in the container:
> >
> > 1470 43 0:32 /rootfs / rw,relatime - aufs none rw,si=cbd79e20de63b17e
> > 1471 1470 0:33 / /proc rw,nosuid,nodev,noexec,relatime - proc proc rw
> > 1472 1470 0:39 / /sys rw,relatime - sysfs sysfs rw
> > 1473 1472 0:18 / /sys/fs/fuse/connections rw,relatime - fusectl none rw
> > 1474 1472 0:6 / /sys/kernel/debug rw,relatime - debugfs none rw
> > 1475 1472 0:10 / /sys/kernel/security rw,relatime - securityfs none rw
> > 1476 1470 0:12 /10 /dev/lxc/console rw,nosuid,noexec,relatime - devpts
> devpts rw,gid=5,mode=620,ptmxmode=000
> > 1477 1470 0:12 /5 /dev/lxc/tty1 rw,nosuid,noexec,relatime - devpts
> devpts rw,gid=5,mode=620,ptmxmode=000
> > 1478 1470 0:12 /6 /dev/lxc/tty2 rw,nosuid,noexec,relatime - devpts
> devpts rw,gid=5,mode=620,ptmxmode=000
> > 1479 1470 0:12 /7 /dev/lxc/tty3 rw,nosuid,noexec,relatime - devpts
> devpts rw,gid=5,mode=620,ptmxmode=000
> > 1480 1470 0:12 /8 /dev/lxc/tty4 rw,nosuid,noexec,relatime - devpts
> devpts rw,gid=5,mode=620,ptmxmode=000
> > 44 1470 0:40 / /dev/pts rw,relatime - devpts devpts
> rw,gid=5,mode=620,ptmxmode=666
> > 50 1470 0:40 /ptmx /dev/ptmx rw,relatime - devpts devpts
> rw,gid=5,mode=620,ptmxmode=666
> > 51 1470 0:41 / /run rw,nosuid,noexec,relatime - tmpfs none
> rw,size=1755388k,mode=755
> > 52 51 0:447 / /run/lock rw,nosuid,nodev,noexec,relatime - tmpfs none
> rw,size=5120k
> > 69 51 0:1087 / /run/shm rw,nosuid,nodev,relatime - tmpfs none rw
> >
> >
> >
> > On Wednesday, 12 March, 2014 at 10:32 pm, Serge Hallyn wrote:
> >
> > > Quoting Dao Quang Minh (dqminh89 at gmail.com (mailto:dqminh89 at gmail.com
> )):
> > > > We havent tested without the bind mounts ( but we can probably try
> it asap
> > > > ). We migrated to 1.0.0 from 0.7.5 about a week ago, and this is the
> first
> > > > time i've seen this bug.
> > > >
> > > > `grep shared /proc/self/mountinfo` doesnt show anything, but
> > > > `/proc/self/mountinfo` does display 2 entries per physical container
> ( i
> > > > guess because of the bind mounts ):
> > > >
> > > > ```
> > > > 1766 39 0:764 / /disk1/container-name rw,relatime - aufs none
> > > > rw,si=70f12540eaf98716
> > > > 1768 38 0:764 / /var/lib/lxc/container-name rw,relatime - aufs none
> > > > rw,si=70f12540eaf98716
> > > >
> > >
> > >
> > > Uh, these show up in the container, right? They should not show up on
> > > the host.
> > >
> > > -serge
> > > _______________________________________________
> > > lxc-users mailing list
> > > lxc-users at lists.linuxcontainers.org (mailto:
> lxc-users at lists.linuxcontainers.org)
> > > http://lists.linuxcontainers.org/listinfo/lxc-users
> > >
> > >
> >
> >
>
> > _______________________________________________
> > lxc-users mailing list
> > lxc-users at lists.linuxcontainers.org
> > http://lists.linuxcontainers.org/listinfo/lxc-users
>
> _______________________________________________
> lxc-users mailing list
> lxc-users at lists.linuxcontainers.org
> http://lists.linuxcontainers.org/listinfo/lxc-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxcontainers.org/pipermail/lxc-users/attachments/20140312/f0402616/attachment.html>