Bug 1317

run please journalctl and check what is spammed in the last part of it.

Created attachment 969 [details]
journalctl -n --lines=100

(In reply to comment #1)
> run please journalctl and check what is spammed in the last part of it.

I have attached the output of
journalctl -n --lines=100

Try 

  systemctl enable avahi-dnsconfd.service

If this help stop systemctl spam?

(In reply to comment #4)
> Try 
> 
>   systemctl enable avahi-dnsconfd.service
> 
> If this help stop systemctl spam?

# systemctl enable avahi-dnsconfd.service
ln -s '/lib/systemd/system/avahi-dnsconfd.service' '/etc/systemd/system/multi-user.target.wants/avahi-dnsconfd.service'


I have no any CPU utilization after that

I also added avahi-dnsconf to the default pacakge list, so next time service should be enabled out of the box.

User has critical CPU utilization again.
Загрузка повторилась. 
    1 root      20   0  5876 3732 2172 R   95  0.2   4:43.74 systemd                  
  414 root      20   0 2135m  39m  38m R   80  2.6   4:06.65 systemd-journal          
 1264 root      20   0 31112 1504 1184 S   18  0.1   0:49.65 rsyslogd  

And he has such error in logs:
В логах куча вот таких записей.

Dec 28 13:53:37 localhost systemd[1]: Failed to start Avahi mDNS/DNS-SD Stack.
Dec 28 13:53:37 localhost systemd[1]: Failed to start Avahi mDNS/DNS-SD Stack.
Dec 28 13:53:37 localhost systemd[1]: Failed to start Avahi mDNS/DNS-SD Stack.
Dec 28 13:53:37 localhost systemd[1]: Failed to start Avahi mDNS/DNS-SD Stack.
Dec 28 13:53:37 localhost systemd[1]: Failed to start Avahi mDNS/DNS-SD Stack.
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Dec 28 13:53:37 localhost systemd[1]: Failed to start Avahi mDNS/DNS-SD Stack.
Dec 28 13:53:37 localhost systemd[1]: Failed to start Avahi mDNS/DNS-SD Stack.

I have such problem on my laptop HP probook 6465b too. All updates are installed.

I have 64bit system.

First - what output for

systemctl | grep failed

Second - (from root)

systemctl mask avahi-daemon.service
systemctl mask avahi-daemon.socket 
systemctl avahi-dnsconfd.service

Is bug appear after that ?


P.S. What reason up importance of bug? It's non-resolved upstream bug. We may only workaround it not resolved anymore in upstream...

Btw, I have this problem on the my laptop too. As a rule it appears after some time I'm working and after I close laptop lid several times (so, the system "sleeps" several times).

As i previously wrote somewhere, the problem is that sometimes avahi-daemon process dies and leaves /var/run/avahi-daemon/pid. Afterwards systemd tries to start avahi-daemon but since pid from previous crashed process is still on it's place it can't be started.

(In reply to comment #10)
> First - what output for
> 
> systemctl | grep failed
> 
> Second - (from root)
> 
> systemctl mask avahi-daemon.service
> systemctl mask avahi-daemon.socket 
> systemctl avahi-dnsconfd.service
> 
> Is bug appear after that ?
> 
> 
> P.S. What reason up importance of bug? It's non-resolved upstream bug. We
> may only workaround it not resolved anymore in upstream...

[root@freshx32 ~]# systemctl | grep failed
atd.service               loaded failed failed        Job spooling tools
atieventsd.service        loaded failed failed        LSB: ATI Events Daemon
cpupower.service          loaded failed failed        Configure CPU power related settings
networkmanager.service    loaded failed failed        LSB: Daemon for automatically switching to best network connection.
systemd-sysctl.service    loaded failed failed        Apply Kernel Variables
systemd-...-setup.service loaded failed failed        Recreate Volatile Files and Directories
winbind.service           loaded failed failed        LSB: Winbind naming service (winbindd)
[root@freshx32 ~]# 

[root@freshx32 ~]# systemctl mask avahi-daemon.service
ln -s '/dev/null' '/etc/systemd/system/avahi-daemon.service'
[root@freshx32 ~]# systemctl mask avahi-daemon.socket 
ln -s '/dev/null' '/etc/systemd/system/avahi-daemon.socket'
[root@freshx32 ~]# systemctl mask avahi-dnsconfd.service
ln -s '/dev/null' '/etc/systemd/system/avahi-dnsconfd.service'

Results I'll see later.

Reason is simple. It is very critical for laptop energy consumption and for all desktop it causes other applications trying to utilize CPU.

My output was written when I have no CPU utilization
systemctl | grep failed

(In reply to comment #10)
> First - what output for
> 
> systemctl | grep failed
> 
> Second - (from root)
> 
> systemctl mask avahi-daemon.service
> systemctl mask avahi-daemon.socket 
> systemctl avahi-dnsconfd.service
> 
> Is bug appear after that ?
> 
> 
> P.S. What reason up importance of bug? It's non-resolved upstream bug. We
> may only workaround it not resolved anymore in upstream...
User has said that problem is present but in does not appear every time.

As for me, the problem is solved by this trick but I had small of the time to check this. If the problem will appear again I'll let you know.

*** Bug 1389 has been marked as a duplicate of this bug. ***

*** Bug 1360 has been marked as a duplicate of this bug. ***

If someone will notice how to reproduce the bug - please do not hesitate to share the experience :)
For now i can suppose that 
ExecStartPre=/bin/rm -f /var/run/avahi-daemon/pid
may help

I suspect that cupsd is the origin of the explosion process systemd and systemd-journald. I'm not sure, but when I turned off my printer, the problem is solved itself instantly!

Actually this helps a lot, because there is indeed such a bug.

On my x586 test system the bug is now (after last updates) shown at every second reboot.
You can do experiments :-)

Ok. I applied a test patch that may help you.

1) Please remove /var/run/avahi-daemon/pid file

2) Check that systemd still consumes much CPU. It should stop doing it after removing /var/run/avahi-daemon/pid

3) Add one this containers to your system ( so that is corresponds to you system arch )
i586
http://abf.rosalinux.ru/downloads/grendizer_personal/container/avahi-897588/RPMS/
x86_64
http://abf.rosalinux.ru/downloads/grendizer_personal/container/avahi-897589/RPMS/

4) Update from it, reboot and then try to reproduce the error one more time.

One workday, four reboot: OK.

OK, i am waiting for feedback until the evening and then i'll push the update to the main.

6 Reboot: - bug appeared again :-(

I also have this problem. If I start Rosalinux without network spot awailable, Systemd starts eating CPU, so I must switch to console and stop avahi service and avahi-daemon.service and avahi-daemon.socket.

Did you try patched avahi,  i linked above ?

The link is broken (i586).

Unfortunaly, same situation with patched Avahi from container.

same problem for me :
Jan 30 21:13:45 localhost systemd[1]: Starting Avahi mDNS/DNS-SD Stack...
Jan 30 21:13:45 localhost systemd[1]: Failed to start Avahi mDNS/DNS-SD Stack.

Does anybody notice that it may happen due to cups ? Do you have cups started on your machine ? Does this happen if you restart cups.service on your PC ?

It is not related to CUPS. I have not cups installed, but I have the issue.

I masked Cups from starting but still have this issue. When it happens only 'systemctl stop avahi-daemon.socket' helps.

fixed by avahi workaround, already published to the repo