devkit-disks-da, gvfs-gdu-volume and dbus-daemon have abnormally high CPU usage

[O. Emmerson]

Binary package hint: devicekit-disks

For the last couple of days devkit-disks-da, gvfs-gdu-volume and dbus-daemon have all had CPU of over 10%. udevd, update-notifier and gdu-notification also have CPU of around 5%.

Googling I found the below threads on ubuntuforums.org
http://ubuntuforums.org/showthread.php?t=1309874
http://ubuntuforums.org/showthread.php?p=8243464

These link show other users are having the problems but they link the issues to ntfs or external media partitions in fstab. I do have ntfs partitions on the drive of the machine I am experiencing the bug with but none of them are in fstab.

ProblemType: Bug
Architecture: i386
Date: Thu Nov 12 23:16:17 2009
DistroRelease: Ubuntu 9.10
ExecutablePath: /usr/lib/devicekit-disks/devkit-disks-daemon
InstallationMedia: Ubuntu 9.10 "Karmic Koala" - Release i386 (20091028.5)
MachineType: LENOVO 423323G
Package: devicekit-disks 007-2ubuntu3
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.31-14-generic root=UUID=c385268b-b9c9-4ebc-9bae-c2080330d660 ro quiet splash
ProcEnviron:

ProcVersionSignature: Ubuntu 2.6.31-14.48-generic
SourcePackage: devicekit-disks
Uname: Linux 2.6.31-14-generic i686
dmi.bios.date: 09/22/2008
dmi.bios.vendor: LENOVO
dmi.bios.version: 6GET11WW(V1.02)
dmi.board.name: KIWDX
dmi.board.vendor: LENOVO
dmi.board.version: REFERENCE
dmi.chassis.asset.tag: No Asset Tag
dmi.chassis.type: 10
dmi.chassis.vendor: No Enclosure
dmi.chassis.version: N/A
dmi.modalias: dmi:bvnLENOVO:bvr6GET11WW(V1.02):bd09/22/2008:svnLENOVO:pn423323G:pvrLenovo3000N500:rvnLENOVO:rnKIWDX:rvrREFERENCE:cvnNoEnclosure:ct10:cvrN/A:
dmi.product.name: 423323G
dmi.product.version: Lenovo 3000 N500
dmi.sys.vendor: LENOVO

[Jarek Dylag (jarek-dylag)]

it is udev related issue described in
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/379780

[Maurizio (maurizio.g)]

Hi, I am affected from this bug too!

I want to comment that I am a little doubtful that this is a duplicate of #481626 for a number of reasons:

1) stopping the udevd daemon doesn't stop the CPU consumption at all
2) killing hald doesn't stop the CPU consumption
3) killing dbus-daemon from messageb user makes the computer reboot
4) I already have a quite recent kernel version:

Linux bastialaptop 2.6.31-15-generic #50-Ubuntu SMP Tue Nov 10 14:53:52 UTC 2009 x86_64 GNU/Linux

I attach the output of some tools to add information... do you need more?

[Maurizio (maurizio.g)]

Sorry for sending all these attachments... hopefully they will be useful!

[Maurizio (maurizio.g)]

I am even more convinced that this bug affects devicekit-disks and is not caused by udev

Today, I was able to let the pc turn on normally, by renaming org.freedesktop.Devicekit.Disks.conf into another file. The system, without finding the config file of DeviceKit-Disks, is not able to start its daemon.
This means that all the operations related to devicekit don't work anymore, but there's no abnormal CPU usage!

cheers

Maurizio

[Maurizio (maurizio.g)]

Hi all... let me add another comment: sometimes now this message starts to flow into the terminal when I make use of dialogs related to nautilus (like, open file):

(gnome-appearance-properties:10278): GVFS-RemoteVolumeMonitor-WARNING **: New owner :1.6621 for volume monitor org.gtk.Private.GduVolumeMonitor connected to the bus; seeding drives/volumes/mounts

(gnome-appearance-properties:10278): GVFS-RemoteVolumeMonitor-WARNING **: invoking List() failed for type GProxyVolumeMonitorGdu: org.freedesktop.DBus.Error.NoReply: Message did not receive a reply (timeout by message bus)

In the terminal, this flow starts with the New owner number increasing constantly... what does this mean??

[BartSchaefer (barton-schaefer)]

I am also experiencing this issue.

Per the tail of this forum thread: http://ubuntuforums.org/showthread.php?p=8110630
I was able to stop the continuous disk access by restarting the udev service.

Here's "top" at first login after a reboot:

  PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
 3305 root 20 0 5988 3744 2240 S 8 0.2 1:06.80 devkit-disks-da
  880 messageb 20 0 3280 1608 748 S 8 0.1 0:56.49 dbus-daemon
 3320 schaefer 20 0 6496 3040 2380 S 5 0.1 0:39.04 gvfs-gdu-volume
    1 root 20 0 2640 1556 1120 S 4 0.1 0:31.39 init
  504 root 18 -2 2372 848 312 S 4 0.0 0:32.33 udevd
 3287 schaefer 20 0 27780 11m 8676 S 4 0.6 0:26.37 update-notifier
 3283 schaefer 20 0 17344 6576 5152 S 3 0.3 0:20.43 gdu-notificatio
  483 root 20 0 2584 1220 944 S 2 0.1 0:11.55 mountall
  498 root 20 0 2148 748 568 S 1 0.0 0:09.24 upstart-udev-br
  500 root 16 -4 2704 1144 300 S 1 0.1 0:05.92 udevd
 1126 root 20 0 92532 20m 9368 S 1 1.0 0:08.95 Xorg

Here it is after "sudo restart udev":

  PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
 1126 root 20 0 108m 23m 9.9m S 3 1.2 0:22.10 Xorg
 7388 schaefer 20 0 241m 81m 24m S 2 4.1 0:10.36 firefox
 7386 schaefer 20 0 2468 1180 884 R 1 0.1 0:00.42 top
 3215 schaefer 20 0 83560 26m 7804 S 0 1.3 0:04.30 compiz.real
 3217 schaefer 20 0 60676 19m 13m S 0 1.0 0:04.29 gnome-panel
    1 root 20 0 2640 1556 1120 S 0 0.1 0:42.54 init
    2 root 15 -5 0 0 0 S 0 0.0 0:00.00 kthreadd
    3 root RT -5 0 0 0 S 0 0.0 0:00.08 migration/0
    4 root 15 -5 0 0 0 S 0 0.0 0:00.01 ksoftirqd/0
    5 root RT -5 0 0 0 S 0 0.0 0:00.00 watchdog/0
    6 root RT -5 0 0 0 S 0 0.0 0:00.01 migration/1

A check with "ps" indicates that devkit-disks-da did NOT exit when udev was restarted, but did cease consuming large amounts of CPU and spinning the disk. There may be intersecting bugs here.

[BartSchaefer (barton-schaefer)]

Oh, forgot -- comment #9 refers to Ubuntu 9.10 running on a Micro-Star International MSI Notebook PR600 (MSI-1637 motherboard, Intel Core2 Duo T7500).

Linux itlan.zanshin.com 2.6.31-17-generic #54-Ubuntu SMP Thu Dec 10 16:20:31 UTC 2009 i686 GNU/Linux

[Maurizio (maurizio.g)]

restarting udev doesn't solve my problem at all.

the only thing I can do is:
1. killall gnome-panel and killall nautilus from the TTY1 terminal at the boot of the pc
2. to give (using htop) a SIGHUP to the /bin/dbus-daemon process, and try not to use any program that is likely to use dbus again (like I can just surf the web). every time it comes back again eating resources, I have to do this again

[Maurizio (maurizio.g)]

let me add the output of Nautilus, to help understanding. I start to think I am affected from a slightly different bug, since I don't see any issue in udev in my system. Should I open another bug? Can I bring all the comments and stuff already uploaded here? Thanks

maurizio@bastialaptop:~$ killall nautilus
maurizio@bastialaptop:~$ nautilus &
[1] 4473
maurizio@bastialaptop:~$
(nautilus:4473): Eel-CRITICAL **: eel_preferences_get_boolean: assertion `preferences_is_initialized ()' failed

** (nautilus:4473): WARNING **: No marshaller for signature of signal 'UploadFinished'

** (nautilus:4473): WARNING **: No marshaller for signature of signal 'DownloadFinished'

** (nautilus:4473): WARNING **: No marshaller for signature of signal 'ShareCreateError'
Initializing nautilus-gdu extension

(nautilus:4473): GVFS-RemoteVolumeMonitor-WARNING **: invoking List() failed for type GProxyVolumeMonitorGdu: org.freedesktop.DBus.Error.NoReply: Message did not receive a reply (timeout by message bus)
Initializing nautilus-dropbox 0.6.1
Nautilus-Share-Message: Called "net usershare info" but it failed: «net usershare» ha restituito l'errore 255: net usershare: cannot open usershare directory /var/lib/samba/usershares. Error Nessun file o directory
Please ask your system administrator to enable user sharing.

[Maurizio (maurizio.g)]

hi all!!

after yesterday's update, the issue is automagically solved on my computer!! :)

[Jiri Lebl (jiri-lebl)]

I have all the updates applied here and the problem is still happening. I have seen the problem using both gnome and xfce. Restarting udev seems to solve the problem until it starts again.

[Lars Kruse (devel-sumpfralle)]

I have the same problem.

For me it was solved by removing the locale file of gvfs - e.g.:
 cd /usr/share/locale/de/LC_MESSAGES
 mv gvfs.mo gvfs.mo.orig
(use YOUR locale instead of "de", of course)

The constantly restarting processes died immediately.

As far as I understand the problem, dbus seems to close connections containing non-ascii characters. These are triggered by umlauts like in the string "Festkörper..." in german locales.

Bug #379780 (marked as a duplicate of this bug) seems to relate to the cd-drive. Thus I doubt, that it is the same problem.

[jy_moustache (jy-poilleux)]

Hi all

Same problem with me.
Did not find any working solutions.

jy

[jy_moustache (jy-poilleux)]

Hi all
So I realized I forgot to give some info here.

I'm running Ubuntu 9.10 Karmic Koala.
I've checked : the bug appears with kernel 2.6.31-9-rt and 2.6.31-19, the last generic kernel avaible to this date.
It appears with gnome and xfce.

I tried restarting udev but it didn't change anything.
I have'nt still tested all the solutions suggested in this thread yet.
I have a NTFS partition which is automatically mounted at startup. I suppressed the corresponding line in /etc/fstab but the problem still occured.

The only thing that worked was killing both devkit-disks-da and gvfs-gdu-volume processes and restarting the computer. Once restarted a had a low CPU usage when computer was not being used. But it worked only once andI haven't been able to reproduce this ever since.

I agree with Lars Kruse (comment #15), I'm not sure that Bug #379780 (marked as a duplicate of this bug) is the same problem. This ones seems to be caused by udev and doesn't have the same symptoms at all.

Thanks

jy

[RickRichardson (rick-richardson-gmail)]

I had this problem. I fixed it by running :

sudo service udev start

I was intending to then 'service stop' ip after I had 'service start'ed it, but the problem went away.

[jy_moustache (jy-poilleux)]

Hi all

I did some more tests. I runned the "sudo stop udev" tip and it did work this time : no more abusive CPU usage. I've been able to reproduce this several times.

I'll be following bug #379780 from now on.

Thanks

jy

[Czollli (czollli)]

Hi all,

I can also confirm the "sudo stop udev" tip!!!
I use:
Linux ubuntu 2.6.31-20-generic #58-Ubuntu SMP Fri Mar 12 05:23:09 UTC 2010 i686 GNU/Linux
on Ubuntu 9.10

Thanks for the investigation of this workaround, high CPU usage was very annoying.

Czollli