[SOLVED] Centos 7 would not shutdown under Proxmox 6.1

[eth]

Hello!

I've just upgraded to proxmox 6.1 and installed the latest Centos 7 template (centos-7-default_20190926_amd64.tar.xz).
When I enter the LXC and try to shut the Centos 7 down, it just won't.


ps aux shows:
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND root 1 0.0 0.9 125304 5236 ? Ss 22:19 0:00 /sbin/init root 37 0.0 1.5 39092 8208 ? Ss 22:19 0:00 /usr/lib/systemd/systemd-journald root 3091 0.0 0.8 116400 4632 ? Ss 22:22 0:00 /bin/bash root 3109 0.0 0.7 155352 4000 ? R+ 22:23 0:00 ps aux x

journalctl -xe shows that all services are stopped, yet the system is still running.

dmesg shows:
[ 5435.948097] audit: type=1400 audit(1577646285.401:82): apparmor="DENIED" operation="mount" info="failed flags match" error=-13 profile="/usr/bin/lxc-start" name="/proc/sys/kernel/random/boot_id" pid=18677 comm="lxc-start" srcname="/dev/.lxc-boot-id" flags="rw, bind" [ 5436.237344] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready [ 5436.237412] vmbr0: port 2(veth200i0) entered blocking state [ 5436.237414] vmbr0: port 2(veth200i0) entered forwarding state

 

[eth]

I created another blank container with the same template and it shuts down perfectly. Probably has something to do with my installed programs within the original container, but it's weird, since "ps aux" shows almost nothing running left.

I wonder if it has something to do with
[ 6295.072255] audit: type=1400 audit(1577647144.515:97): apparmor="DENIED" operation="mount" info="failed flags match" error=-13 profile="/usr/bin/lxc-start" name="/proc/sys/kernel/random/boot_id" pid=3000 comm="lxc-start" srcname="/dev/.lxc-boot-id" flags="rw, bind"

 

[eth]

When I do "strace shutdown -P now", I see:

recvmsg(4, {msg_name(0)=NULL, msg_iov(1)=[{"l\4\1\1!\2\0\0007\0\0\0\236\0\0\0\1\1o\0008\0\0\0", 24}], msg_controllen=32, [{cmsg_len=28, cmsg_level=SOL_SOCKET, cmsg_type=SCM_CREDENTIALS, {pid=1, uid=0, gid=0}}], msg_flags=MSG_CMSG_CLOEXEC}, MSG_DONTWAIT|MSG_NOSIGNAL|MSG_CMSG_CLOEXEC) = 24 recvmsg(4, {msg_name(0)=NULL, msg_iov(1)=[{"/org/freedesktop/systemd1/unit/p"..., 697}], msg_controllen=32, [{cmsg_len=28, cmsg_level=SOL_SOCKET, cmsg_type=SCM_CREDENTIALS, {pid=1, uid=0, gid=0}}], msg_flags=MSG_CMSG_CLOEXEC}, MSG_DONTWAIT|MSG_NOSIGNAL|MSG_CMSG_CLOEXEC) = 697 recvmsg(4, {msg_name(0)=NULL, msg_iov(1)=[{"l\4\1\1\364\2\0\0008\0\0\0\236\0\0\0\1\1o\0008\0\0\0", 24}], msg_controllen=32, [{cmsg_len=28, cmsg_level=SOL_SOCKET, cmsg_type=SCM_CREDENTIALS, {pid=1, uid=0, gid=0}}], msg_flags=MSG_CMSG_CLOEXEC}, MSG_DONTWAIT|MSG_NOSIGNAL|MSG_CMSG_CLOEXEC) = 24 recvmsg(4, {msg_name(0)=NULL, msg_iov(1)=[{"/org/freedesktop/systemd1/unit/p"..., 908}], msg_controllen=32, [{cmsg_len=28, cmsg_level=SOL_SOCKET, cmsg_type=SCM_CREDENTIALS, {pid=1, uid=0, gid=0}}], msg_flags=MSG_CMSG_CLOEXEC}, MSG_DONTWAIT|MSG_NOSIGNAL|MSG_CMSG_CLOEXEC) = 908 recvmsg(4, 0x7ffe8b2982f0, MSG_DONTWAIT|MSG_NOSIGNAL|MSG_CMSG_CLOEXEC) = -1 EAGAIN (Resource temporarily unavailable) ppoll([{fd=4, events=POLLIN}], 1, NULL, NULL, 8

but I can do ctrl-c and the system is still running.

 

[eth]

Ok, so I traced it down to some of my custom /etc/init.d/ scripts which looks like are conflicting with the latest centos 7.
Sorry, guys.

 

[shantanu]

Sometimes I observe this behavior with running VMs/CTs. Everything works OK once I reboot the host machine.