meta data for this page
  •  

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
vm:proxmox:lxc:issues [2021/03/22 15:44] – niziakvm:proxmox:lxc:issues [2023/07/28 12:13] (current) – niziak
Line 1: Line 1:
 ====== LXC Issues ====== ====== LXC Issues ======
 +
 +===== lxc_init: Failed to run lxc.hook.pre-start for container =====
 +
 +After upgrade guest system from Debian 12.0 to 12.1.
 +
 +<code bash>
 +lxc-start -lDEBUG -o error.log -F -n <ContainerID>
 +</code>
 +
 +<code>unsupported debian version '12.1'</code>
 +
 +PVE (''pve-container'') needs upgrade.
 +
 +
 +===== apply caps: operation not permitted: unknown. =====
 +
 +BalenaOS Build inside privileged LXC:
 +<code>
 +docker: Error response from daemon: OCI runtime create failed: container_linux.go:380: starting container process caused: apply caps: operation not permitted: unknown.
 +</code>
 +
 +Solution (not secure!):
 +<file container.conf>
 +lxc.apparmor.profile: unconfined
 +lxc.cgroup.devices.allow: a
 +lxc.cap.drop:
 +</file>
 +
 +Source: [[https://danthesalmon.com/running-docker-on-proxmox/]]
 +
 +
 +===== Slow login into container =====
 +
 +see below
 +
 +===== Failed at step NAMESPACE spawning /lib/systemd/systemd-logind: Permission denied =====
 +
 +Debian Bullseye in unprivileged container:
 +
 +<code>
 +systemd[579]: systemd-logind.service: Failed to set up mount namespacing: /run/systemd/unit-root/proc: Permission denied
 +systemd[579]: systemd-logind.service: Failed at step NAMESPACE spawning /lib/systemd/systemd-logind: Permission denied
 +</code>
 +SOLUTION: enable container nesting.
 +
 +
 +
 +===== cannot stop container =====
 +Container works, responds to pings but it is not possible to SSH or attach.
 +
 +Normal commands to stop or reboot doesn't help (even ''lxc-stop -k'').
 +
 +**CAUSE:** Container was freezed for snapshot. All processess are in 'D' state. Cannot be killed.
 +**SOLUTION:**
 +<code bash>
 +echo THAWED > /sys/fs/cgroup/freezer/lxc/200/freezer.state
 +</code>
 +
 +**Info**: 
 +  * [[https://www.kernel.org/doc/Documentation/cgroup-v1/freezer-subsystem.txt|Freezer subsystem]]
 +  * [[https://www.kernel.org/doc/html/latest/admin-guide/cgroup-v1/freezer-subsystem.html]]
 +  * [[https://www.kernel.org/doc/html/latest//power/freezing-of-tasks.html|Freezing of tasks]]
 +
 +
 +==== Investigation ====
 +
 +So killing container is solution:
 +<code bash>
 +pstree -p
 +
 +           ├─lxc-start(3747487)───systemd(3747514)─┬─agetty(3748048)
 +           │                                       ├─agetty(3748049)
 +
 +kill -9 3747514
 +</code>
 +
 +
 +Now it is not possible to start LXC container again. Debugging:
 +<code bash>
 +lxc-start -o lxc-start.log -lDEBUG -F -n 200
 +cat lxc-start.log
 +
 +lxc-start 200 20210325085035.665 INFO     conf - conf.c:run_script_argv:331 - Executing script "/usr/share/lxc/hooks/lxc-pve-prestart-hook" for container "200", config section "lxc"
 +lxc-start 200 20210325085036.126 DEBUG    conf - conf.c:run_buffer:303 - Script exec /usr/share/lxc/hooks/lxc-pve-prestart-hook 200 lxc pre-start produced output: failed to remove directory '/sys/fs/cgroup/net_cls/lxc/200/ns': Device or resource busy
 +</code>
 +
 +**Reason:** after killing container systemd, orphaned cgroups left.
 +<code bash>
 +find /sys/fs/cgroup/*/lxc/200 -depth -type d -print -delete
 +
 +# Lot of errors:
 +find: cannot delete ‘/sys/fs/cgroup/freezer/lxc/200’: Device or resource busy
 +</code>
 +
 +All processess from container 200 are in 'D' state ''Uninterruptible Sleep''
 +<code bash>
 +ps axl | awk '$10 ~ /D/'
 +</code>
 +
 +<code bash>
 +echo w > /proc/sysrq_trigger
 +
 +[587314.999001] smbd            D    0 1181293  42630 0x00004184
 +[587314.999002] Call Trace:
 +[587314.999004]  __schedule+0x2e6/0x6f0
 +[587314.999005]  schedule+0x33/0xa0
 +[587314.999007]  __refrigerator+0x44/0x160
 +[587314.999009]  get_signal+0x814/0x850
 +[587314.999011]  do_signal+0x34/0x6e0
 +[587314.999013]  ? wait_woken+0x80/0x80
 +[587314.999014]  ? __audit_syscall_exit+0x236/0x290
 +[587314.999016]  exit_to_usermode_loop+0x90/0x130
 +[587314.999018]  do_syscall_64+0x160/0x190
 +[587314.999020]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
 +
 +</code>
 +
 +So it looks like whole container cgroup was freezed for snapshot and problem happens.
 +
  
 ===== nested docker in cpulimit ===== ===== nested docker in cpulimit =====