Bug#802306: packagekitd eating up RAM
Hi Matthias, On Sat, 13 Feb 2016 16:33:05 +0100, Matthias Klumpp wrote: > Can you check if this issue still appears with the latest PackageKit > in unstable (v1.1.0)? > I fixed a few minor memory leaks that I found, and there are also a > few other changes which might have an impact on this - although I am > still a bit in the dark on why this issue happens at all (still can't > reproduce it :-/). Sorry for the lack of response, I haven't had the time to look into this in detail. But it looks like the memory leak is fixed, I have a two-week-old packagekitd process only using 33MB of RAM. Regards, Stephen pgpAJIkclTj3R.pgp Description: OpenPGP digital signature
Bug#802306: packagekitd eating up RAM
Hi! Can you check if this issue still appears with the latest PackageKit in unstable (v1.1.0)? I fixed a few minor memory leaks that I found, and there are also a few other changes which might have an impact on this - although I am still a bit in the dark on why this issue happens at all (still can't reproduce it :-/). Cheers, Matthias
Bug#802306: packagekitd eating up RAM
Package: packagekit Version: 1.0.11-2 Followup-For: Bug #802306 -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Dear Maintainer, This still exists in 1.0.11-2. - -- System Information: Debian Release: stretch/sid APT prefers unstable APT policy: (500, 'unstable'), (500, 'stable'), (1, 'experimental') Architecture: amd64 (x86_64) Kernel: Linux 4.0.0 (SMP w/4 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages packagekit depends on: ii gdebi-core 0.9.5.7 ii libapt-inst2.0 1.2.2 ii libapt-pkg5.0 1.2.2 ii libc6 2.21-7 ii libgcc1 1:5.3.1-8 ii libglib2.0-02.46.2-3 ii libglib2.0-bin 2.46.2-3 ii libgstreamer1.0-0 1.6.3-1 ii libpackagekit-glib2-18 1.0.11-2 ii libpolkit-gobject-1-0 0.105-14.1 ii libsqlite3-03.10.2-1 ii libstdc++6 5.3.1-8 ii libsystemd0 228-6 ii policykit-1 0.105-14.1 Versions of packages packagekit recommends: ii packagekit-tools 1.0.11-2 Versions of packages packagekit suggests: ii appstream [appstream-index] 0.9.1-1 - -- no debconf information -BEGIN PGP SIGNATURE- Version: GnuPG v1 iF4EAREIAAYFAla9860ACgkQa46zoGXPuqkl3gD/ftsReHe6141R7xjorrDIz0tC oKEFwBL6+Ar/T/mQ8k0A+wahaUzu84JP4Y/8TyCLf9xwOxb/PV0o9kfcrBQ7158M =aDWV -END PGP SIGNATURE-
Bug#802306: packagekitd eating up RAM
Package: packagekit Version: 1.0.10-1 Followup-For: Bug #802306 Me too, Currently packagekit is eating about 6GB of memory. I ran strings on the core file and the following bunch keeps repeating: -- WwanHardwareEnabled WimaxEnabled WimaxHardwareEnabled ActiveConnections PrimaryConnection WirelessHardwareEnabled Connectivity Devices DBus WirelessEnabled WirelessHardwareEnabled ActivatingConnection DBus NetworkingEnabled /org/freedesktop/NetworkManager/Devices/0 /org/freedesktop/NetworkManager/Devices/0 DBus edesktop% Metered provided% PrimaryConnectionType /org/freedesktop/NetworkManager/Devices/1 /org/freedesktop/NetworkManager/Devices/2 /org/freedesktop/NetworkManager/Devices/3 /org/freedesktop/NetworkManager/Devices/4 -- I have statically configured interfaces (/etc/network/interfaces) and two qemu virtual machines. All connected to br0. No WLAN/WWAN/Bluetooth adapters in system. % nmcli d DEVICE TYPE STATE CONNECTION vnet0 tap disconnected -- vnet1 tap disconnected -- br0 bridgeunmanaged -- eth0ethernet unmanaged -- lo loopback unmanaged -- -- System Information: Debian Release: stretch/sid APT prefers testing APT policy: (900, 'testing'), (300, 'unstable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.2.0-1-amd64 (SMP w/8 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages packagekit depends on: ii gdebi-core 0.9.5.7 ii libapt-inst1.7 1.0.10.2 ii libapt-pkg4.16 1.0.10.2 ii libc6 2.19-22 ii libgcc1 1:5.2.1-23 ii libglib2.0-02.46.2-1 ii libglib2.0-bin 2.46.2-1 ii libgstreamer1.0-0 1.6.1-1 ii libpackagekit-glib2-18 1.0.10-1 ii libpolkit-gobject-1-0 0.105-14 ii libsqlite3-03.9.2-1 ii libstdc++6 5.2.1-23 ii libsystemd0 228-2 ii policykit-1 0.105-14 Versions of packages packagekit recommends: ii packagekit-tools 1.0.10-1 Versions of packages packagekit suggests: pn appstream-index -- no debconf information
Bug#802306: packagekitd eating up RAM
Package: packagekit Version: 1.0.11-1 Followup-For: Bug #802306 Dear Maintainer, * What led up to the situation? * What exactly did you do (or not do) that was effective (or ineffective)? * What was the outcome of this action? * What outcome did you expect instead? *** End of the template - remove these template lines *** I confirm that this bug is present in an up to date Debian testing installation (stretch). I am running Debian testing via SparkyLinux, with Openbox desktop. A few days ago I noticed that there were 3 packaged processes each consuming 25% of total RAM (4 GB). I stopped the packaged service via systemctl, system performance improved. Two days ago I rebooted for a kernel upgrade, packaged restarted and is now consuming 3x 17 = 51% of RAM. RAM usage is increasing noticeably from day to day. -- System Information: Debian Release: stretch/sid APT prefers testing APT policy: (500, 'testing') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.3.0-1-amd64 (SMP w/2 CPU cores) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages packagekit depends on: ii gdebi-core 0.9.5.7 ii libapt-inst2.0 1.1.10 ii libapt-pkg5.0 1.1.10 ii libc6 2.21-6 ii libgcc1 1:5.3.1-5 ii libglib2.0-02.46.2-3 ii libglib2.0-bin 2.46.2-3 ii libgstreamer1.0-0 1.6.2-1 ii libpackagekit-glib2-18 1.0.11-1 ii libpolkit-gobject-1-0 0.105-14 ii libsqlite3-03.9.2-1 ii libstdc++6 5.3.1-5 ii libsystemd0 228-2+b1 ii policykit-1 0.105-14 Versions of packages packagekit recommends: ii packagekit-tools 1.0.11-1 Versions of packages packagekit suggests: pn appstream-index -- no debconf information -- systemctl status: ● packagekit.service - PackageKit Daemon Loaded: loaded (/lib/systemd/system/packagekit.service; static; vendor preset: enabled) Active: active (running) since Sat 2016-01-09 20:24:20 CET; 1 day 19h ago Main PID: 1854 (packagekitd) CGroup: /system.slice/packagekit.service └─1854 /usr/lib/packagekit/packagekitd Jan 09 20:24:20 kirtley systemd[1]: Starting PackageKit Daemon... Jan 09 20:24:20 kirtley PackageKit[1854]: daemon start Jan 09 20:24:20 kirtley systemd[1]: Started PackageKit Daemon.
Bug#802306: packagekitd eating up RAM
Does this problem also appear when running PK on the command-line? Please install the packagekit-dbg package and valgrind, then kill the running packagekitd daemon and run the following command in a terminal session: sudo G_DEBUG=gc-friendly G_SLICE=always-malloc valgrind --tool=memcheck --leak-check=full --leak-resolution=high --xml=yes --xml-file=vgdump.xml --num-callers=20 /usr/lib/packagekit/packagekitd --verbose Then please send me the vgdump.xml file (attached to the bug) - you can send it in a private mail, if you want to. You may want to perform some tasks on PK via pkcon, like running `pkcon refresh` to give the daemon some work to do. Please quit the daemon by issueing the following DBus call: gdbus call --system \ --dest org.freedesktop.PackageKit \ --object-path /org/freedesktop/PackageKit \ --method org.freedesktop.PackageKit.SuggestDaemonQuit Cheers, Matthias
Bug#802306: packagekitd eating up RAM
On Tue, 5 Jan 2016 17:11:41 +0100, Matthias Klumpp wrote: > Hmm, are you doing anything special with PK which could trigger this > issue, so I can reproduce it? No, like Sam I'm just running Gnome, without even using gnome-software. > PK itself should terminate after idling for a while, freeing any > resources it still uses. It's packagekitd that's using tons of RAM, isn't that a service started by systemd? "systemctl status packagekit" produces ● packagekit.service - PackageKit Daemon Loaded: loaded (/lib/systemd/system/packagekit.service; static; vendor preset: enabled) Active: active (running) since Mon 2016-01-04 20:04:22 CET; 24h ago Main PID: 3863 (packagekitd) CGroup: /system.slice/packagekit.service └─3863 /usr/lib/packagekit/packagekitd Jan 04 20:04:22 heffalump.sk2.org systemd[1]: Starting PackageKit Daemon... Jan 04 20:04:22 heffalump.sk2.org PackageKit[3863]: daemon start Jan 04 20:04:22 heffalump.sk2.org systemd[1]: Started PackageKit Daemon. on my system just now (it was booted a day ago). > In order to debug a potential memory leak, I would first need to be > able to reproduce it... I was thinking of running it with valgrind... Regards, Stephen pgp6uVNleuM_W.pgp Description: OpenPGP digital signature
Bug#802306: packagekitd eating up RAM
On Tue, Jan 05, 2016, Matthias Klumpp wrote: > Hmm, are you doing anything special with PK which could trigger this > issue, so I can reproduce it? No, I’m afraid that’s the strange thing, I don’t do anything with PK. I run a GNOME desktop but have never launched software or performed any action related to packagekitd as far as I can tell. I only install stuff through apt-get. I do have /usr/bin/gnome-software running, which seems to be related to PK, but I suppose it comes with modern GNOME. I do have a large sources.list, though not insanely so (see attachment). strace()ing the PK process shows that it constantly checks for the existence of /system-update and /var/lib/PackageKit/prepared-update. The comment in pk_offline_auth_invalidate() suggests that this should be a rare occurrence, yet it happens every second (I could misunderstand the code; it’s my first time having a look at it). It’s pretty hard to troubleshoot anything because the box is currently quite unresponsive, but I can try to come up with more information if it can help. > PK itself should terminate after idling for a while, freeing any > resources it still uses. Well, it’d be nice if it released those gigabytes of RAM, sure, but I’d be even happier if it actually didn’t need several gigabytes of RAM in order to work :-) Regards, -- Sam. # Current deb [arch=amd64,i386] http://ftp.fr.debian.org/debian/ experimental main non-free contrib deb-src http://ftp.fr.debian.org/debian/ experimental main non-free contrib deb [arch=amd64,i386] http://ftp.fr.debian.org/debian/ sid main non-free contrib deb-src http://ftp.fr.debian.org/debian/ sid main non-free contrib # Releases deb [arch=amd64,i386] http://ftp.fr.debian.org/debian/ stretch main non-free contrib deb-src http://ftp.fr.debian.org/debian/ stretch main non-free contrib deb [arch=amd64,i386] http://ftp.fr.debian.org/debian/ jessie main non-free contrib deb-src http://ftp.fr.debian.org/debian/ jessie main non-free contrib deb [arch=amd64,i386] http://ftp.fr.debian.org/debian/ squeeze main non-free contrib deb-src http://ftp.fr.debian.org/debian/ squeeze main non-free contrib deb [arch=amd64,i386] http://ftp.fr.debian.org/debian/ wheezy main non-free contrib deb-src http://ftp.fr.debian.org/debian/ wheezy main non-free contrib # Updates deb [arch=amd64,i386] http://security.debian.org/ jessie/updates main contrib non-free deb-src http://security.debian.org/ jessie/updates main contrib non-free deb [arch=amd64,i386] http://security.debian.org/ squeeze/updates main contrib non-free deb-src http://security.debian.org/ squeeze/updates main contrib non-free deb [arch=amd64,i386] http://security.debian.org/ wheezy/updates main contrib non-free deb-src http://security.debian.org/ wheezy/updates main contrib non-free
Bug#802306: packagekitd eating up RAM
Hmm, are you doing anything special with PK which could trigger this issue, so I can reproduce it? PK itself should terminate after idling for a while, freeing any resources it still uses. In order to debug a potential memory leak, I would first need to be able to reproduce it... Cheers, Matthias
Bug#802306: packagekitd eating up RAM
On Sat, Dec 19, 2015, Stephen Kitt wrote: > Dear Maintainer, > > I'm also seeing excessive RAM use; packagekitd is currently using > 4.8GB of RAM (on my 32GB system). I’m tempted to suggest raising the severity of this bug. It has become unbearable, the insane memory usage of packagekitd causes processes to be randomly killed, compilations to fail, X to hang in inconsistent states (the mouse loses its ability to focus windows), several of our machines regularly become unusable for several minutes due to this apparent leak. Sometimes it’s faster to reboot the machine than to wait for it to become responsible again. Just as I am writing this, packagekitd is using 1.8 GiB (RSS). /proc indicates that most of it is heap memory. This really feels like a serious bug to me. Regards, -- Sam.
Bug#802306: packagekitd eating up RAM
Package: packagekit Version: 1.0.11-1 Followup-For: Bug #802306 Dear Maintainer, I'm also seeing excessive RAM use; packagekitd is currently using 4.8GB of RAM (on my 32GB system). Regards, Stephen -- System Information: Debian Release: stretch/sid APT prefers testing APT policy: (500, 'testing'), (500, 'stable'), (200, 'unstable'), (1, 'experimental') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.2.0-1-amd64 (SMP w/8 CPU cores) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages packagekit depends on: ii gdebi-core 0.9.5.7 ii libapt-inst2.0 1.1.4 ii libapt-pkg5.0 1.1.4 ii libc6 2.21-4 ii libgcc1 1:5.2.1-23 ii libglib2.0-02.46.2-1 ii libglib2.0-bin 2.46.2-1 ii libgstreamer1.0-0 1.6.1-1 ii libpackagekit-glib2-18 1.0.11-1 ii libpolkit-gobject-1-0 0.105-14 ii libsqlite3-03.9.2-1 ii libstdc++6 5.2.1-23 ii libsystemd0 228-2 ii policykit-1 0.105-14 Versions of packages packagekit recommends: ii packagekit-tools 1.0.11-1 Versions of packages packagekit suggests: ii appstream [appstream-index] 0.8.6-1 -- no debconf information
Bug#802306: packagekitd eating up RAM
Control: reassign -1 packagekit 1.0.10-1 On Lu, 19 oct 15, 09:24:01, Mike Gabriel wrote: > Package: packagekitd > Severity: important > Version: 1.0.10-1 > > Hi, > > on my Debian sid testing I have been observing high memory consumptions for > a while caused by packagekitd. > > I have to kill packagekitd every once in a while, being a bit fed up with > that now, I have removed it completely from the system. > > > Greets, > Mike > -- > > DAS-NETZWERKTEAM > mike gabriel, herweg 7, 24357 fleckeby > fon: +49 (1520) 1976 148 > > GnuPG Key ID 0x25771B31 > mail: mike.gabr...@das-netzwerkteam.de, http://das-netzwerkteam.de > > freeBusy: > https://mail.das-netzwerkteam.de/freebusy/m.gabriel%40das-netzwerkteam.de.xfb -- http://wiki.debian.org/FAQsFromDebianUser Offtopic discussions among Debian users and developers: http://lists.alioth.debian.org/mailman/listinfo/d-community-offtopic http://nuvreauspam.ro/gpg-transition.txt signature.asc Description: Digital signature
Bug#802306: packagekitd eating up RAM
Package: packagekitd Severity: important Version: 1.0.10-1 Hi, on my Debian sid testing I have been observing high memory consumptions for a while caused by packagekitd. I have to kill packagekitd every once in a while, being a bit fed up with that now, I have removed it completely from the system. Greets, Mike -- DAS-NETZWERKTEAM mike gabriel, herweg 7, 24357 fleckeby fon: +49 (1520) 1976 148 GnuPG Key ID 0x25771B31 mail: mike.gabr...@das-netzwerkteam.de, http://das-netzwerkteam.de freeBusy: https://mail.das-netzwerkteam.de/freebusy/m.gabriel%40das-netzwerkteam.de.xfb pgpPXkiRS_VzL.pgp Description: Digitale PGP-Signatur