On Tue, Mar 24, 2015 at 04:40:00PM +0200, Ruslan Kuprieiev wrote: > Hi Tycho, > > On 03/24/2015 04:21 PM, Tycho Andersen wrote: > >Hi Ruslan, > > > >On Tue, Mar 24, 2015 at 10:14:02AM +0200, Ruslan Kuprieiev wrote: > >>Hi Axel, > >> > >>On 03/24/2015 01:17 AM, Axel Schöner wrote: > >>>Am Montag, 23. März 2015, 15:42:47 schrieb Ruslan Kuprieiev: > >>>>Hi! > >>>> > >>>>I'm trying to use lxc-checkpoint to dump/restore simple lxc container. > >>>> > >>>>I added: > >>>>|# hax for criu > >>>> > >>>>lxc.console = none > >>>>lxc.tty = 0 > >>>>lxc.cgroup.devices.deny = c 5:1 rwm| > >>>> > >>>>to container config, but when I'm trying to run > >>>> > >>>>|lxc-checkpoint -n $name -D $checkpoint_dir -s -v| > >>>> > >>>>I get: > >>>>Error (mount.c:624): 113: ./sys/fs/cgroup/perf_event doesn't have a > >>>>proper root mount > >>>The problem is, that "/sys/fs/cgroup/perf_event" is in your case a > >>>bindmount > >>>from outside (host). > >>>CRIU can dump extern ressources which are bind mounted, you can find > >>>details > >>>on: http://www.criu.org/External_bind_mounts. > >>>lxc-checkpoint does not care about /sys/fs/cgroup/perf_event. > >>Thanks for a hint. > >> > >>But shouldn't lxc-checkpoint just work? I mean, it should dump/restore lxc > >>container, so it probably should handle perf_event too. > >Unfortunately, the recent autodev patches have broken lxc-checkpoint. > >I'm working on some patches to criu to properly resolve mount points as > >autodev does them, but I haven't gotten anything working yet. > > > >Another problem is that systemd mounts propagate differently (SLAVE vs > >PRIVATE), and that confuses criu as well. I'm now back focused more on > >migration/criu vs LXD proper, so fixes should be rolling in in the > >coming weeks, hopefully. > > > Oh! So everything should be working on a bit older ubuntu, right? > I saw your doom migration demo and your instructions in your blog, > so looks like it clearly worked well for you =).
Yep :). For that I used utopic hosts with utopic containers with no LXCFS. LXC hash 0080bebf3195 is probably a good one to use. You might be able to work around the autodev stuff with: lxc.mount.auto = lxc.mount.auto = proc:rw sys:rw > I might just try using a previous one, as I'm currently using the > most recent ubuntu server for my experiments. > > >That said, can you show /proc/<pid>/mountinfo for init? It looks like > >your cgroups may be mounted in yet another configuration that we're > >not prepared to handle that I haven't seen before. > > Sure. Do you mean init like the init process of the host, right? If so, here > it is: Sorry, I meant the init process of the container; something like: sudo cat /proc/$(sudo lxc-info -n $container -H -i)/mountinfo Thanks, Tycho > 16 21 0:15 / /sys rw,nosuid,nodev,noexec,relatime - sysfs sysfs rw > 17 21 0:3 / /proc rw,nosuid,nodev,noexec,relatime - proc proc rw > 18 21 0:5 / /dev rw,relatime - devtmpfs udev > rw,size=239064k,nr_inodes=59766,mode=755 > 19 18 0:12 / /dev/pts rw,nosuid,noexec,relatime - devpts devpts > rw,gid=5,mode=620,ptmxmode=000 > 20 21 0:16 / /run rw,nosuid,noexec,relatime - tmpfs tmpfs > rw,size=50132k,mode=755 > 21 0 252:0 / / rw,relatime - ext4 /dev/mapper/ubuntu--vg-root > rw,errors=remount-ro,data=ordered > 22 16 0:17 / /sys/fs/cgroup rw,relatime - tmpfs none rw,size=4k,mode=755 > 23 16 0:18 / /sys/fs/fuse/connections rw,relatime - fusectl none rw > 24 16 0:6 / /sys/kernel/debug rw,relatime - debugfs none rw > 25 16 0:10 / /sys/kernel/security rw,relatime - securityfs none rw > 26 20 0:19 / /run/lock rw,nosuid,nodev,noexec,relatime - tmpfs none > rw,size=5120k > 27 20 0:20 / /run/shm rw,nosuid,nodev,relatime - tmpfs none rw > 28 20 0:21 / /run/user rw,nosuid,nodev,noexec,relatime - tmpfs none > rw,size=102400k,mode=755 > 29 16 0:22 / /sys/fs/pstore rw,relatime - pstore none rw > 57 21 0:35 / /var/lib/lxcfs rw,nosuid,nodev,relatime - fuse.lxcfs lxcfs > rw,user_id=0,group_id=0,allow_other > 58 21 8:1 / /boot rw,relatime - ext2 /dev/sda1 rw > 59 22 0:34 / /sys/fs/cgroup/systemd rw,nosuid,nodev,noexec,relatime - cgroup > systemd > rw,release_agent=/run/cgmanager/agents/cgm-release-agent.systemd,name=systemd > > Thanks, > Ruslan > > >Thanks, > > > >Tycho > > > >>> > >>> > >>>>Could you give me a hint on what is going on? > >>>>I also saw that other people do experience the same problem: > >>>>http://tycho.ws/blog/2014/09/container-migration.html > >>>> > >>>>Thanks, > >>>>Ruslan > >>Thanks, > >>Ruslan > >>_______________________________________________ > >>lxc-devel mailing list > >>lxc-devel@lists.linuxcontainers.org > >>http://lists.linuxcontainers.org/listinfo/lxc-devel > >_______________________________________________ > >lxc-devel mailing list > >lxc-devel@lists.linuxcontainers.org > >http://lists.linuxcontainers.org/listinfo/lxc-devel > _______________________________________________ lxc-devel mailing list lxc-devel@lists.linuxcontainers.org http://lists.linuxcontainers.org/listinfo/lxc-devel
