1317 2012-12-24 14:50:29 +0400 systemd occupies 100% utilization of CPU 2026-04-17 23:40:37 +0300 1 1 2 ROSA-based products ROSA Fresh Packages from Main Fresh All Linux RESOLVED FIXED High major --- 1 firstlevel bugs alex.burmashev alexander.kazantsev alexander.petryakov altadeos denis.koryavov kostiagol mtzseb r0g3r shura0 v.potapov --- systemd-194-10-rosa2012.1.i586 known oldest_to_newest 6063 0 960 firstlevel 2012-12-24 14:50:29 +0400 Created attachment 960 systemctl-a.txt Description of problem: systemd occupies 100% utilization of CPU I have such information of the top output PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 1 root 20 0 5876 3416 2000 R 97 0.2 55:12.21 systemd 2875 root 20 0 2226m 33m 32m R 88 1.6 45:58.23 systemd-journal # uname -r 3.6.10-nrj-desktop-1rosa I have attached the output of systemctl -a Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. 6064 1 alex.burmashev 2012-12-24 14:51:37 +0400 run please journalctl and check what is spammed in the last part of it. 6140 2 969 firstlevel 2012-12-25 08:59:56 +0400 Created attachment 969 journalctl -n --lines=100 6141 3 firstlevel 2012-12-25 09:00:15 +0400 (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 6143 4 alexander.kazantsev 2012-12-25 09:06:47 +0400 Try systemctl enable avahi-dnsconfd.service If this help stop systemctl spam? 6182 5 firstlevel 2012-12-25 13:07:56 +0400 (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 6191 6 alex.burmashev 2012-12-25 15:21:37 +0400 I also added avahi-dnsconf to the default pacakge list, so next time service should be enabled out of the box. 6380 7 firstlevel 2012-12-28 09:19:18 +0400 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. 6435 8 firstlevel 2012-12-29 09:33:50 +0400 I have such problem on my laptop HP probook 6465b too. All updates are installed. 6436 9 firstlevel 2012-12-29 09:34:23 +0400 I have 64bit system. 6437 10 alexander.kazantsev 2012-12-29 10:18:16 +0400 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... 6440 11 denis.koryavov 2012-12-29 12:59:46 +0400 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). 6443 12 alex.burmashev 2012-12-29 13:29:11 +0400 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. 6461 13 firstlevel 2012-12-29 15:25:49 +0400 (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. 6463 14 firstlevel 2012-12-29 15:28:58 +0400 My output was written when I have no CPU utilization systemctl | grep failed 6499 15 firstlevel 2012-12-30 19:43:35 +0400 (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. 6522 16 denis.koryavov 2013-01-03 21:12:32 +0400 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. 6524 17 denis.koryavov 2013-01-03 21:14:28 +0400 *** Bug 1389 has been marked as a duplicate of this bug. *** 6526 18 denis.koryavov 2013-01-03 21:15:20 +0400 *** Bug 1360 has been marked as a duplicate of this bug. *** 6604 19 alex.burmashev 2013-01-09 14:33:56 +0400 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 6676 20 mtzseb 2013-01-10 18:21:12 +0400 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! 6677 21 alex.burmashev 2013-01-10 18:24:50 +0400 Actually this helps a lot, because there is indeed such a bug. 6798 22 v.potapov 2013-01-14 15:44:07 +0400 On my x586 test system the bug is now (after last updates) shown at every second reboot. You can do experiments :-) 6811 23 alex.burmashev 2013-01-14 17:58:02 +0400 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. 6873 24 v.potapov 2013-01-15 17:00:33 +0400 One workday, four reboot: OK. 6912 25 alex.burmashev 2013-01-16 14:01:39 +0400 OK, i am waiting for feedback until the evening and then i'll push the update to the main. 6913 26 v.potapov 2013-01-16 14:23:54 +0400 6 Reboot: - bug appeared again :-( 7111 27 kostiagol 2013-01-25 14:07:10 +0400 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. 7112 28 alex.burmashev 2013-01-25 14:21:42 +0400 Did you try patched avahi, i linked above ? 7113 29 kostiagol 2013-01-25 14:25:43 +0400 The link is broken (i586). 7127 30 r0g3r 2013-01-26 02:08:24 +0400 Unfortunaly, same situation with patched Avahi from container. 7314 31 mtzseb 2013-01-31 00:16:16 +0400 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. 7339 32 alex.burmashev 2013-01-31 14:16:12 +0400 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 ? 7775 33 shura0 2013-02-10 13:25:29 +0400 It is not related to CUPS. I have not cups installed, but I have the issue. 7776 34 kostiagol 2013-02-10 13:36:51 +0400 I masked Cups from starting but still have this issue. When it happens only 'systemctl stop avahi-daemon.socket' helps. 8394 35 alex.burmashev 2013-03-01 14:01:40 +0400 fixed by avahi workaround, already published to the repo 87664 36 pastordidi 2026-04-17 23:39:21 +0300 The content of attachment 960 has been deleted 87673 37 pastordidi 2026-04-17 23:40:37 +0300 The content of attachment 969 has been deleted 960 2012-12-24 14:50:29 +0400 2012-12-24 14:50:29 +0400 systemctl-a.txt systemctl-a.txt text/plain 0 firstlevel 969 2012-12-25 08:59:56 +0400 2012-12-25 08:59:56 +0400 journalctl -n --lines=100 journalctl.txt text/plain 0 firstlevel