[systemd-devel] Container restart issue: Failed to attach 1 to compat systemd cgroup

[Lewis Gaul]

Hi all,

I've come across an issue when restarting a systemd container, which I'm
seeing on a CentOS 8.2 VM but not able to reproduce on an Ubuntu 20.04 VM
(both cgroups v1).

The failure looks as follows, hitting the warning condition at
https://github.com/systemd/systemd/blob/v245/src/shared/cgroup-setup.c#L279:

[root at localhost ubuntu-systemd]# podman run -it --privileged --name ubuntu
--detach ubuntu-systemd
5e4ab2a36681c092f4ef937cf03b25a8d3d7b2fa530559bf4dac4079c84d0313

[root at localhost ubuntu-systemd]# podman restart ubuntu
5e4ab2a36681c092f4ef937cf03b25a8d3d7b2fa530559bf4dac4079c84d0313

[root at localhost ubuntu-systemd]# podman logs ubuntu | grep -B6 -A2 'Set
hostname'
systemd 245.4-4ubuntu3.19 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 podman.
Detected architecture x86-64.

Welcome to Ubuntu 20.04.5 LTS!

Set hostname to <5e4ab2a36681>.
[  OK  ] Created slice system-getty.slice.
[  OK  ] Created slice system-modprobe.slice.
--
systemd 245.4-4ubuntu3.19 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 podman.
Detected architecture x86-64.

Welcome to Ubuntu 20.04.5 LTS!

Set hostname to <5e4ab2a36681>.

*Failed to attach 1 to compat systemd cgroup
/machine.slice/libpod-5e4ab2a36681c092f4ef937cf03b25a8d3d7b2fa530559bf4dac4079c84d0313.scope/init.scope:
No such file or directory*[  OK  ] Created slice system-getty.slice.


If using docker instead of podman (still on CentOS 8.2) the container
actually exits after restart (when hitting the code at
https://github.com/systemd/systemd/blob/v245/src/core/cgroup.c#L2972):

[root at localhost ubuntu-systemd]# docker logs ubuntu | grep -C5 'Set
hostname'
Detected virtualization docker.
Detected architecture x86-64.

Welcome to Ubuntu 20.04.5 LTS!

Set hostname to <523caa1f03e9>.
[  OK  ] Created slice system-getty.slice.
[  OK  ] Created slice system-modprobe.slice.
[  OK  ] Created slice User and Session Slice.
[  OK  ] Started Dispatch Password Requests to Console Directory Watch.
[  OK  ] Started Forward Password Requests to Wall Directory Watch.
--
Detected virtualization docker.
Detected architecture x86-64.

Welcome to Ubuntu 20.04.5 LTS!

Set hostname to <523caa1f03e9>.




*Failed to attach 1 to compat systemd cgroup
/system.slice/docker-523caa1f03e9c96a6a12a55fb07df995c6e4b3a27e18585cbeda869b943ae728.scope/init.scope:
No such file or directoryFailed to open pin file: No such file or
directoryFailed to allocate manager object: No such file or
directory[!!!!!!] Failed to allocate manager object.Exiting PID 1...*


Does anyone know what might be causing this? Is it a systemd bug? I can
copy the info into a GitHub issue if that's helpful.

Thanks,
Lewis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/systemd-devel/attachments/20230109/3d1e4e07/attachment.htm>