[CRIU] dumping LXC 1.0 container failed - unable to open irmap-cache: No such file or directory

[David Shwatrz]

Hi,
Thanks, Andrew, for your quick response.
The container is fedora 20 (as well as the host).

It is impossible to unmount systemd cgroups in this container:

For example,
umount /sys/fs/cgroup/systemd
umount: /sys/fs/cgroup/systemd: target is busy
        (In some cases useful info about processes that
         use the device is found by lsof(8) or fuser(1).)

I will test with an Ubuntu container.

Following here is the requested output of cat /proc/4983/mountinfo
gives:
255 195 8:5 /var/lib/lxc/fedora/rootfs / rw,relatime master:1 - ext4
/dev/sda5 rw,data=ordered
256 255 0:5 /.lxc/fedora.ef29892550ef9cbf /dev rw,nosuid master:2 -
devtmpfs devtmpfs rw,size=1514576k,nr_inodes=378644,mode=755
196 256 0:39 / /dev/pts rw,relatime - devpts devpts
rw,gid=5,mode=620,ptmxmode=666
193 255 0:37 / /proc rw,nosuid,nodev,noexec,relatime - proc proc rw
194 255 0:40 / /sys rw,nosuid,nodev,noexec,relatime - sysfs sysfs rw
197 256 0:41 / /dev/shm rw,nosuid,nodev - tmpfs tmpfs rw
198 255 0:42 / /run rw,nosuid,nodev - tmpfs tmpfs rw,mode=755
199 194 0:43 / /sys/fs/cgroup rw,nosuid,nodev,noexec - tmpfs tmpfs rw,mode=755
200 199 0:20 / /sys/fs/cgroup/systemd rw,nosuid,nodev,noexec,relatime
- cgroup cgroup
rw,xattr,release_agent=/usr/lib/systemd/systemd-cgroups-agent,name=systemd
201 199 0:22 / /sys/fs/cgroup/cpuset rw,nosuid,nodev,noexec,relatime -
cgroup cgroup rw,cpuset,clone_children
202 199 0:23 / /sys/fs/cgroup/cpu,cpuacct
rw,nosuid,nodev,noexec,relatime - cgroup cgroup rw,cpuacct,cpu
203 199 0:24 / /sys/fs/cgroup/memory rw,nosuid,nodev,noexec,relatime -
cgroup cgroup rw,memory
204 199 0:25 / /sys/fs/cgroup/devices rw,nosuid,nodev,noexec,relatime
- cgroup cgroup rw,devices
205 199 0:26 / /sys/fs/cgroup/freezer rw,nosuid,nodev,noexec,relatime
- cgroup cgroup rw,freezer
206 199 0:27 / /sys/fs/cgroup/net_cls rw,nosuid,nodev,noexec,relatime
- cgroup cgroup rw,net_cls
207 199 0:28 / /sys/fs/cgroup/blkio rw,nosuid,nodev,noexec,relatime -
cgroup cgroup rw,blkio
208 199 0:29 / /sys/fs/cgroup/perf_event
rw,nosuid,nodev,noexec,relatime - cgroup cgroup rw,perf_event
209 199 0:30 / /sys/fs/cgroup/hugetlb rw,nosuid,nodev,noexec,relatime
- cgroup cgroup rw,hugetlb
210 194 0:33 / /sys/kernel/config rw,relatime - configfs configfs rw
211 194 0:6 / /sys/kernel/debug rw,relatime - debugfs debugfs rw
212 255 0:44 / /tmp rw - tmpfs tmpfs rw
213 256 0:36 / /dev/mqueue rw,relatime - mqueue mqueue rw
214 256 0:45 / /dev/hugepages rw,relatime - hugetlbfs hugetlbfs rw


DavidS

On Wed, Mar 5, 2014 at 4:45 PM, Andrew Vagin <avagin at parallels.com> wrote:
> On Wed, Mar 05, 2014 at 03:54:33PM +0200, David Shwatrz wrote:
>> Christopher,
>> Yes, you are absolutely right. In tests which I had done with a simple
>> process (not  in a containers environment), everything with c/r worked
>> fine, but this same message still appeared.
>>
>> >Error (mount.c:430): Mount 327 (master_id: 2 shared_id: 0) has >unreachable sharing
>>
>> Any ideas how can we tell to which mount/mount point is this message related ?
>
> Could you show content of /proc/4983/mountinfo?
>
> What distribution do you use in CT?
>
> criu doesn't support cgroup, hugetlbfs, fusectl, configfs, mqueue,
> debugfs file systems in containers, so you should umount them from CT.
>
> http://criu.org/LXC
>
>>
>> Regards.
>> DavidS
>>
>>
>>
>> On Wed, Mar 5, 2014 at 2:55 PM, Christopher Covington
>> <cov at codeaurora.org> wrote:
>> > On 03/05/2014 05:32 AM, David Shwatrz wrote:
>> >> Hello,
>> >> I am trying to checkpoint an LXC 1.0 container and it fails.
>> >> I am trying it with crtools 1.2.
>> >>
>> >> The command I use is:
>> >> criu dump -t 4983 -v4 -D images --shell-job --ext-unix-sk -o dump.log
>> >> --tcp-established --file-locks
>> >> and the log file shows this as a first error:
>> >> unable to open irmap-cache: No such file or directory
>> >> the full log fine is in:
>> >> http://pastebin.com/e3MKxmBg
>> >
>> > The irmap stuff is spurious. This looks like your real error:
>> >
>> > Error (mount.c:430): Mount 327 (master_id: 2 shared_id: 0) has unreachable sharing
>> >
>> > Regards,
>> > Christopher
>> >
>> > --
>> > Employee of Qualcomm Innovation Center, Inc.
>> > Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
>> > hosted by the Linux Foundation.
>> _______________________________________________
>> CRIU mailing list
>> CRIU at openvz.org
>> https://lists.openvz.org/mailman/listinfo/criu