[systemd mkinitcpio ] New weird messages during shutdown after upgrade

Potomac · 2013-12-19 13:46:47

after upgrading mkinitcpio ( 16 ) and systemd ( 208-3 ) I notice new messages during shutdown, the shutdown is a little bit slower now :

http://hfr-rehost.dev.syn.fr/self/e300a … 747a4f.png

is it possible to avoid these multiple lines "unmounting /oldroot" and "cgroup : option or name mismatch, new: 0x0"", old: 0x4 "systemd" ?

It seems that something has changed in mkinitcpio package,

this change creates a random bug in virtualbox with my archlinux 32 bits virtual machine, during shutdown systemd can waits almost 1 minute after this message :

"[Ok] Started Generate shutdown-ramfs"

if I downgrade mkinitcpio it solves the problem

-- mod edit: read the Forum Etiquette and only post thumbnails http://wiki.archlinux.org/index.php/For … s_and_Code [jwr] --

Last edited by Potomac (2013-12-19 13:47:28)

stqn · 2013-12-19 14:37:19

All I can do is point you to this mail: https://mailman.archlinux.org/pipermail … 25742.html

brain0 · 2013-12-19 19:08:40

Potomac wrote:

after upgrading mkinitcpio ( 16 ) and systemd ( 208-3 ) I notice new messages during shutdown, the shutdown is a little bit slower now :
http://hfr-rehost.dev.syn.fr/self/e300a … 747a4f.png
is it possible to avoid these multiple lines "unmounting /oldroot" and "cgroup : option or name mismatch, new: 0x0"", old: 0x4 "systemd" ?
It seems that something has changed in mkinitcpio package,

These messages indicate a clean umount of all file systems. If you don't want a clean shutdown, you can mask the service

systemctl mask mkinitcpio-generate-shutdown-ramfs.service

Potomac wrote:

this change creates a random bug in virtualbox with my archlinux 32 bits virtual machine, during shutdown systemd can waits almost 1 minute after this message

I pretty much doubt that is related at all. I get random hangs during shutdown all the time, just every 10th shutdown or so. No idea what's going on there.

Potomac · 2013-12-19 20:22:43

it's related because if I mask the service "mkinitcpio-generate-shutdown-ramfs.service" the random bug is gone under virtualbox,

doing a "Generate shutdown-ramfs" during shutdown on a virtual machine seems non safe, maybe this action triggers a new bug in virtualbox

brain0 · 2013-12-19 22:19:45

Potomac wrote:

it's related because if I mask the service "mkinitcpio-generate-shutdown-ramfs.service" the random bug is gone under virtualbox,

How do you know that? Personally, I can't reproduce this at all, it just happens every 10th shutdown or so, entirely independent of anything else I do (or so it seems).

Potomac wrote:

doing a "Generate shutdown-ramfs" during shutdown on a virtual machine seems non safe, maybe this action triggers a new bug in virtualbox

This sentence is pure nonsense.

/etc/hellacog · 2014-01-12 14:53:50

I am having the same issue in VMware Fusion 5.0.4. I also notice that the VM seems to periodically have a problem mounting the SWAP volume at startup. Could this be related? It intermittently seems to hang at "[Ok] Started Generate shutdown-ramfs" for as long as 3 minutes before completing the shutdown. If someone could provide some useful tips to get rig of this it would be greatly appreciated. No "beat-downs" please...

brain0 · 2014-01-13 22:16:36

As I said, this is unrelated.

[Ok] Started Generate shutdown-ramfs

This means that it actually finished running the generation. It's hanging somewhere else.

drankinatty · 2014-01-18 00:32:33

brain0 wrote:

As I said, this is unrelated.
[Ok] Started Generate shutdown-ramfs
This means that it actually finished running the generation. It's hanging somewhere else.

What can we do to find out what is causing it? I got hit by this for the first time on Jan 16:

Jan 16 12:43:20 phoinix systemd[1]: Stopped MariaDB database server.
Jan 16 12:44:03 phoinix systemd[1]: Started Generate shutdown-ramfs.
Jan 16 12:44:09 phoinix systemd[1]: Started Generate shutdown-ramfs.
Jan 16 12:44:19 phoinix systemd[1]: Started Generate shutdown-ramfs.
Jan 16 12:44:20 phoinix systemd[1]: Started Generate shutdown-ramfs.
Jan 16 12:44:22 phoinix systemd[1]: Started Generate shutdown-ramfs.
Jan 16 12:44:23 phoinix systemd[1]: Started Generate shutdown-ramfs.
Jan 16 12:44:24 phoinix systemd[1]: Started Generate shutdown-ramfs.
Jan 16 12:44:26 phoinix systemd[1]: Started Generate shutdown-ramfs.
Jan 16 12:44:36 phoinix systemd[1]: Started Generate shutdown-ramfs.
Jan 16 12:44:38 phoinix systemd[1]: Started Generate shutdown-ramfs.
-- Reboot --

Where it says -- Reboot -- I had to do a hard-reset to reboot it. This occurred updating linux (3.12.6-1 -> 3.12.7-2). Is there anything I can provide.

brain0 · 2014-01-18 09:11:02

In your case, it keeps starting the service over and over - this doesn't make sense to me, since it should only do so once. I don't know how to even reproduce this. Sure, you can

# systemctl mask mkinitcpio-generate-shutdown-ramfs.service

but even if it helps, it still doesn't tell us what is wrong.

drankinatty · 2014-01-18 09:42:34

The box is at a remote site. I'll drop by and reboot it and see if it hangs tomorrow. That is what worried me. I usually do updates and reboots remotely. I was lucky I was there when it happened. I had no problems rebooting with kernels before 3.12.6-1. I'm hoping this was a one time deal. I'll let you know tomorrow.

chetwisniewski · 2014-03-03 02:20:03

I am getting this as well on a Xen DOM0.

Last thing it does is say this then hang until I remove the power and restart.

 Mar 02 17:59:08 adamwest.zuzax.local systemd[1]: Started Generate shutdown-ramfs.
Mar 02 17:59:10 adamwest.zuzax.local systemd[1]: Started Generate shutdown-ramfs.
Mar 02 17:59:11 adamwest.zuzax.local systemd[1]: Started Generate shutdown-ramfs.
Mar 02 17:59:11 adamwest.zuzax.local systemd[1]: Started Generate shutdown-ramfs.
Mar 02 17:59:11 adamwest.zuzax.local systemd[1]: Started Generate shutdown-ramfs.
Mar 02 17:59:11 adamwest.zuzax.local systemd[1]: Started Generate shutdown-ramfs.
Mar 02 17:59:11 adamwest.zuzax.local systemd[1]: Started Generate shutdown-ramfs.

Is there a way to debug the service to see what the problem is?

Chester

brain0 · 2014-03-03 11:11:29

You could increase the log level of systemd (systemd.log_level=debug on the command line). Then more details will be available in the journal, hopefully.

Arch Linux

#1 2013-12-19 13:46:47

[systemd mkinitcpio ] New weird messages during shutdown after upgrade

#2 2013-12-19 14:37:19

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

#3 2013-12-19 19:08:40

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

#4 2013-12-19 20:22:43

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

#5 2013-12-19 22:19:45

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

#6 2014-01-12 14:53:50

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

#7 2014-01-13 22:16:36

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

#8 2014-01-18 00:32:33

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

#9 2014-01-18 09:11:02

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

#10 2014-01-18 09:42:34

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

#11 2014-03-03 02:20:03

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

#12 2014-03-03 11:11:29

Re: [systemd mkinitcpio ] New weird messages during shutdown after upgrade

Board footer