But from the output it shows that is mounting a fake cgroupfs, no?
Have exactly the same issue after upgrade from 3.21 to 3.22 (on arm64). Tried both versions from snap (rev 13718 3.22/stable and rev 13764 3.22/candidate).
Inside containers I see only 1 process init:
root@ns01:~# ps aux USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND root 1 0.0 0.0 12116 3068 ? Ss 14:13 0:00 /sbin/init root 205 0.0 0.0 7300 3364 ? Ss 14:13 0:00 bash root 215 0.0 0.0 8680 2764 ? R+ 14:13 0:00 ps aux
systemd 237 running in system mode. (+PAM +AUDIT +SELINUX +IMA +APPARMOR +SMACK +SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ +LZ4 +SECCOMP +BLKID +ELFUTILS +KMOD -IDN2 +IDN -PCRE2 default-hierarchy=hybrid) Detected virtualization lxc. Detected architecture arm64. Welcome to Ubuntu 18.04.1 LTS! Set hostname to <ns01>. Failed to create compat systemd cgroup /lxc.payload/ns01/init.scope: Invalid argument Failed to attach 1 to compat systemd cgroup /lxc.payload/ns01/init.scope: No such file or directory Failed to open pin file: No such file or directory Failed to allocate manager object: No such file or directory [ESC[0;1;31m!!!!!!ESC[0m] Failed to allocate manager object, freezing. Freezing execution.
Set to “security.privileged” and server reboot didn’t help.
We’ve reproduced the issue on a local test system.
@brauner will investigate shortly.
Tentative fix merged in lxcfs and being included in the candidate snap now.
We’ll validate on CentOS 7 before pushing to stable later today.
Seems to work fine here. Just need to wait for arm64 to be done building before releasing to stable.
Rolled out to stable.
Confirmed, arm64 version
stable: 3.22 2020-03-16 (13771) 62MB -
works now without issues.
Thanks for fix that.
Good to hear, sorry about that.
Thanks for the fix.