Page 1 of 4 123 ... LastLast
Results 1 to 10 of 34

Thread: Why is my hexacore Xeon at 100% for copying some files via samba?

  1. #1

    Default Why is my hexacore Xeon at 100% for copying some files via samba?

    Hello again!

    I have the following setup


    Code:
    NAS 1 (Raspberrry Pi 2, Raspbian Buster, Samba version 4.9.5-Debian) ---SMB--- Dolphin Workstation (Xeon Hexacore, Tumbleweed KDE) ---SMB--- NAS 2 (Odroid XU4, Debian Stretch, Samba version 4.5.16-Debian)
    In the Dolphin on the TW machine (Samba version 4.13.0-git.138.ff2d5480c67SUSE-oS15.5-x86_64) I started copying over some 500 GB of files from NAS 1 to NAS 2 (actually to put some load on the NAS 2 machine, which is newly configured).

    Interestingly, the ONLY machine under heavy load in this funny experiment is the hexacore Xeon machine running TW, which is at nearly 100% load (according to htop load average: 11.92 10.95 10.95, temps around 80°C), while the small machines in the NAS are idle (htop on Odroid (octacore) 0.04, on Raspberry (quadcore) 0.22).

    Can somebody enlighten me, why the samba on the TW has 9 (nine) processes

    https://paste.opensuse.org/d73a8102

    usinge the whole machine, while the ARM SOCs are nearly doing nothing... ?


    PS: On both NAS there are mdadm RAID1 configs with WDred HDDs (2x 1TB on NAS 1, 2x 4TB on NAS 2)
    Kind regards

    raspu

  2. #2
    Join Date
    Jun 2008
    Location
    Auckland, NZ
    Posts
    22,081
    Blog Entries
    1

    Default Re: Why is my hexacore Xeon at 100% for copying some files via samba?

    Multiple connections created to handle the file transfers perhaps? I think that is normal behaviour. More info here...
    https://wiki.samba.org/index.php/Sam..._process_model

    The 'max smbd processes' parameter can be used to limit the number of smbd processes if necessary. Refer...
    Code:
    man smb.conf
    openSUSE Leap 15.2; KDE Plasma 5

  3. #3

    Default Re: Why is my hexacore Xeon at 100% for copying some files via samba?

    Hi and thanks!

    I have no problem with 9 processes, I only want to learn: WHAT ARE THESE PROCESSES DOING (100% load on XEON) if there is no load on the ARM source and target machine? ;-)
    Kind regards

    raspu

  4. #4

    Default Re: Why is my hexacore Xeon at 100% for copying some files via samba?

    Really? Nobody in the whole wide world knows what this smb is doing whole day long? XEON still under full load here, 65 GB to go...
    Kind regards

    raspu

  5. #5
    Join Date
    Jun 2008
    Location
    Podunk
    Posts
    29,679
    Blog Entries
    15

    Default Re: Why is my hexacore Xeon at 100% for copying some files via samba?

    Quote Originally Posted by suse_rasputin View Post
    Really? Nobody in the whole wide world knows what this smb is doing whole day long? XEON still under full load here, 65 GB to go...
    Hi
    So it's sucking data from one machine to the host (caching and processing) and then sending it off at the same time to the secondary machine? What about iotop? Maybe the GUI manager...?

    Why not just transfer from NAS to NAS?
    Cheers Malcolm °¿° SUSE Knowledge Partner (Linux Counter #276890)
    SUSE SLE, openSUSE Leap/Tumbleweed (x86_64) | GNOME DE
    If you find this post helpful and are logged into the web interface,
    please show your appreciation and click on the star below... Thanks!

  6. #6

    Default Re: Why is my hexacore Xeon at 100% for copying some files via samba?

    Hi and thanks for reply! :-)

    I wanted to put some load on the smb connection, expecting the load on the raspberry pi and maybe on the odroid (which both are idle), but not on the machine running the Dolphin/TW...

    I could have used scp, I know, but I wanted to test smb...

    iotop ist here:

    https://paste.opensuse.org/ac9cc6fb

    anything else?

    I still have 9 smb processes putting some 100% load on all 6 cores, what do they do? :-o
    Kind regards

    raspu

  7. #7
    Join Date
    Jun 2008
    Location
    Podunk
    Posts
    29,679
    Blog Entries
    15

    Default Re: Why is my hexacore Xeon at 100% for copying some files via samba?

    Quote Originally Posted by suse_rasputin View Post
    Hi and thanks for reply! :-)

    I wanted to put some load on the smb connection, expecting the load on the raspberry pi and maybe on the odroid (which both are idle), but not on the machine running the Dolphin/TW...

    I could have used scp, I know, but I wanted to test smb...

    iotop ist here:

    https://paste.opensuse.org/ac9cc6fb

    anything else?

    I still have 9 smb processes putting some 100% load on all 6 cores, what do they do? :-o
    Hi
    Likely those entries kworker/1:0H-kblockd, probably the network interface module on the host....

    Trigger a backtrace and see if anything common shows up....

    Code:
    echo 1 > /proc/sysrq-trigger
    Check dmesg output after above and see if anything interesting pops out.
    Cheers Malcolm °¿° SUSE Knowledge Partner (Linux Counter #276890)
    SUSE SLE, openSUSE Leap/Tumbleweed (x86_64) | GNOME DE
    If you find this post helpful and are logged into the web interface,
    please show your appreciation and click on the star below... Thanks!

  8. #8

    Default Re: Why is my hexacore Xeon at 100% for copying some files via samba?

    I have here:

    Code:
    dell6:~ # echo 1 > /proc/sysrq-trigger
    dell6:~ #
    Sorry, no idea how to "Trigger a backtrace"... ;-)

    in dmesg nothing happenz at all, just some nuvo spam:

    Code:
    ...
    [ 1638.801611] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084 failed with error -2
    [ 1638.801633] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084d failed with error -2
    [ 1638.801636] nouveau 0000:03:00.0: msvld: unable to load firmware data
    [ 1638.801638] nouveau 0000:03:00.0: msvld: init failed, -19
    [22723.741140] perf: interrupt took too long (2502 > 2500), lowering kernel.perf_event_max_sample_rate to 79750
    [41272.793609] perf: interrupt took too long (3132 > 3127), lowering kernel.perf_event_max_sample_rate to 63750
    [64116.006095] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084 failed with error -2
    [64116.006106] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084d failed with error -2
    [64116.006108] nouveau 0000:03:00.0: msvld: unable to load firmware data
    [64116.006109] nouveau 0000:03:00.0: msvld: init failed, -19
    [77278.961998] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084 failed with error -2
    [77278.962008] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084d failed with error -2
    [77278.962009] nouveau 0000:03:00.0: msvld: unable to load firmware data
    [77278.962011] nouveau 0000:03:00.0: msvld: init failed, -19
    [163404.425735] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084 failed with error -2
    [163404.425745] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084d failed with error -2
    [163404.425747] nouveau 0000:03:00.0: msvld: unable to load firmware data
    [163404.425748] nouveau 0000:03:00.0: msvld: init failed, -19
    [246811.739421] sysrq: Changing Loglevel
    [246811.739425] sysrq: Loglevel set to 1
    Kind regards

    raspu

  9. #9
    Join Date
    Jun 2008
    Location
    Podunk
    Posts
    29,679
    Blog Entries
    15

    Default Re: Why is my hexacore Xeon at 100% for copying some files via samba?

    Quote Originally Posted by suse_rasputin View Post
    I have here:

    Code:
    dell6:~ # echo 1 > /proc/sysrq-trigger
    dell6:~ #
    Sorry, no idea how to "Trigger a backtrace"... ;-)

    in dmesg nothing happenz at all, just some nuvo spam:

    Code:
    ...
    [ 1638.801611] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084 failed with error -2
    [ 1638.801633] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084d failed with error -2
    [ 1638.801636] nouveau 0000:03:00.0: msvld: unable to load firmware data
    [ 1638.801638] nouveau 0000:03:00.0: msvld: init failed, -19
    [22723.741140] perf: interrupt took too long (2502 > 2500), lowering kernel.perf_event_max_sample_rate to 79750
    [41272.793609] perf: interrupt took too long (3132 > 3127), lowering kernel.perf_event_max_sample_rate to 63750
    [64116.006095] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084 failed with error -2
    [64116.006106] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084d failed with error -2
    [64116.006108] nouveau 0000:03:00.0: msvld: unable to load firmware data
    [64116.006109] nouveau 0000:03:00.0: msvld: init failed, -19
    [77278.961998] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084 failed with error -2
    [77278.962008] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084d failed with error -2
    [77278.962009] nouveau 0000:03:00.0: msvld: unable to load firmware data
    [77278.962011] nouveau 0000:03:00.0: msvld: init failed, -19
    [163404.425735] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084 failed with error -2
    [163404.425745] nouveau 0000:03:00.0: Direct firmware load for nouveau/nva8_fuc084d failed with error -2
    [163404.425747] nouveau 0000:03:00.0: msvld: unable to load firmware data
    [163404.425748] nouveau 0000:03:00.0: msvld: init failed, -19
    [246811.739421] sysrq: Changing Loglevel
    [246811.739425] sysrq: Loglevel set to 1
    Hi
    OK, open another terminal and run journalctl -f then in the other window run that command, should see backtrace output for every cpu?
    Cheers Malcolm °¿° SUSE Knowledge Partner (Linux Counter #276890)
    SUSE SLE, openSUSE Leap/Tumbleweed (x86_64) | GNOME DE
    If you find this post helpful and are logged into the web interface,
    please show your appreciation and click on the star below... Thanks!

  10. #10

    Default Re: Why is my hexacore Xeon at 100% for copying some files via samba?

    Ok in one terminal I have:

    Code:
     journalctl -f
    Hint: You are currently not seeing messages from other users and the system.
          Users in the 'systemd-journal' group can see all messages. Pass -q to
          turn off this notice.
    -- Logs begin at Sun 2020-07-26 17:31:35 CEST. --
    Sep 30 17:45:11 dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:11 dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:11 dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:14 dell6 plasmashell[2965]: file:///usr/lib64/qt5/qml/QtQuick/Controls/Styles/Plasma/BusyIndicatorStyle.qml:26:9: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:45:14 dell6 systemd[2475]: Started app-org.kde.konsole-b10ed26cdaac41b7a4d5a039c9fe73c5.scope.
    Sep 30 17:45:14  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:17  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:17  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:20  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:20  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:23  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:26  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:29  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:35  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:44  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:47  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:47  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:50  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:53  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:53  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:56  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:45:59  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:02  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:05  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:11  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:11  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:14  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:14  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:17  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:20  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:23  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:29  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:32  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:35  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:36  dell6 plasmashell[2965]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:285:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:46:36  dell6 plasmashell[2965]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:285:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:46:36  dell6 plasmashell[2965]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:285:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:46:36  dell6 plasmashell[2965]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:285:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:46:37  dell6 plasmashell[2965]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:285:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:46:37  dell6 plasmashell[2965]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:285:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:46:37  dell6 plasmashell[2965]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:285:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:46:37  dell6 plasmashell[2965]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:285:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:46:37  dell6 plasmashell[2965]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:285:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:46:38  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:38  dell6 plasmashell[2965]: qml: temp unit: 0
    Sep 30 17:46:39  dell6 plasmashell[2965]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:285:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
    Sep 30 17:46:41  dell6 plasmashell[2965]: qml: temp unit: 0

    while in the other

    Code:
    dell6:~ # echo 1 > /proc/sysrq-trigger
    dell6:~ # echo 1 > /proc/sysrq-trigger
    dell6:~ # echo 1 > /proc/sysrq-trigger
    dell6:~ # echo 1 > /proc/sysrq-trigger
    dell6:~ #
    Makes no difference if I run journalctl -f as root...
    Kind regards

    raspu

Page 1 of 4 123 ... LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •