Mageia forum

Hi folks,

just wanted to report that with the latest updates and kernel, kontact (specifically kmail) in VERY unstable again (it was good for a few weeks).

It now locks up entirely when I delete mail (using imap as the backend) for some reason. It is particularly bad if I delete a single message, but also fails fairly regularly when deleting a block of messages.

I have no idea how to fault find this or if its an issue that should be reported upstream or not (happy to do so if it should be).

Thanks.

Peter.

It's a know issue (or at least be me :p )
I found in fact 2 issue :
- the maildispatcher agent is crashing sometimes (akonadi's related)
- kmail itself crash & i did submit https://bugs.kde.org/show_bug.cgi?id=289892

hum in fact it could be in fact related to mariadb.

pnunn > what is your arch ? i586 or x86_64 ?
could you check that the myslqd launched by akonadi is running ?

arch is x86_64 and ps -A | grep mysqld does return an instance.

Peter.

do you have a backtrace ?

hey, can you also perhaps post your ~/.local/share/akonadi/db_data/mysql.err file (or mysql.err.old)?

No backtrace.. but the file you want is attached.

Thanks.

Peter.

hmm, where can i find this attachment?

OK.. it said it was attached.. let me try again.

OK.. had to change the extension for it to happen. The .old is attached here, if you need the other one I can do that as well.

Thanks.

Pete.

it doesn't show the error yet, i guess you didn't start it again, can you also post your mysql.err file?

also, if possible a system log from that time? like messages or syslog?

Hi Alian,

I had trouble posting last night for some reason (not sure why) so will try again now.

I've just fired it up and its crashed again, so here are (I hope) the logs you were after.

There is nothing in /var/log/messages that I can see

Code: Select all

Dec 28 08:37:48 localhost udisks-daemon[2552]: **** /proc/self/mountinfo changed
Dec 28 08:37:48 localhost udisks-daemon[2552]: **** /proc/self/mountinfo changed
Dec 28 08:38:05 localhost dbus[2561]: [system] Activating via systemd: service name='org.freedesktop.RealtimeKit1' unit='rtkit-daemon.service'
Dec 28 08:38:05 localhost dbus-daemon[2561]: dbus[2561]: [system] Activating via systemd: service name='org.freedesktop.RealtimeKit1' unit='rtkit-daemon.service'
Dec 28 08:38:05 localhost dbus[2561]: [system] Successfully activated service 'org.freedesktop.RealtimeKit1'
Dec 28 08:38:05 localhost dbus-daemon[2561]: dbus[2561]: [system] Successfully activated service 'org.freedesktop.RealtimeKit1'
Dec 28 08:38:05 localhost rtkit-daemon[9566]: Successfully made thread 9565 of process 9565 (/usr/bin/pulseaudio) owned by 'pnunn' high priority at nice level -11.
Dec 28 08:38:06 localhost rtkit-daemon[9566]: Successfully made thread 9569 of process 9565 (/usr/bin/pulseaudio) owned by 'pnunn' RT at priority 5.
Dec 28 08:38:06 localhost rtkit-daemon[9566]: Successfully made thread 9570 of process 9565 (/usr/bin/pulseaudio) owned by 'pnunn' RT at priority 5.
Dec 28 08:38:06 localhost dbus[2561]: [system] Activating via systemd: service name='org.bluez' unit='dbus-org.bluez.service'
Dec 28 08:38:06 localhost dbus-daemon[2561]: dbus[2561]: [system] Activating via systemd: service name='org.bluez' unit='dbus-org.bluez.service'
Dec 28 08:38:06 localhost dbus[2561]: [system] Activation via systemd failed for unit 'dbus-org.bluez.service': Unit dbus-org.bluez.service failed to load: No such file or directory. See system logs and 'systemctl status dbus-org.bluez.service' for details.
Dec 28 08:38:06 localhost dbus-daemon[2561]: dbus[2561]: [system] Activation via systemd failed for unit 'dbus-org.bluez.service': Unit dbus-org.bluez.service failed to load: No such file or directory. See system logs and 'systemctl status dbus-org.bluez.service' for details.
Dec 28 08:38:06 localhost pulseaudio[9565]: [pulseaudio] bluetooth-util.c: Error from ListAdapters reply: org.freedesktop.systemd1.LoadFailed
Dec 28 08:38:08 localhost rtkit-daemon[9566]: Successfully made thread 9620 of process 9620 (/usr/bin/pulseaudio) owned by 'pnunn' high priority at nice level -11.
Dec 28 08:38:08 localhost pulseaudio[9620]: [pulseaudio] pid.c: Daemon already running.
Dec 28 08:38:13 localhost rtkit-daemon[9566]: Successfully made thread 9744 of process 9744 (/usr/bin/pulseaudio) owned by 'pnunn' high priority at nice level -11.
Dec 28 08:38:13 localhost pulseaudio[9744]: [pulseaudio] pid.c: Daemon already running.
Dec 28 08:38:13 localhost rtkit-daemon[9566]: Successfully made thread 9747 of process 9747 (/usr/bin/pulseaudio) owned by 'pnunn' high priority at nice level -11.
Dec 28 08:38:13 localhost pulseaudio[9747]: [pulseaudio] pid.c: Daemon already running.
Dec 28 08:38:13 localhost mgaapplet[9749]: ### Program is starting ###
Dec 28 08:38:16 localhost kernel: martian source 192.168.0.191 from 192.168.0.42, on dev wlan0
Dec 28 08:38:16 localhost kernel: ll header: ff:ff:ff:ff:ff:ff:00:26:b9:b6:fd:94:08:06
Dec 28 08:38:16 localhost net_applet[9763]: ### Program is starting ###
Dec 28 08:38:24 localhost dbus[2561]: [system] Activating service name='org.blueman.Mechanism' (using servicehelper)
Dec 28 08:38:24 localhost dbus-daemon[2561]: dbus[2561]: [system] Activating service name='org.blueman.Mechanism' (using servicehelper)
Dec 28 08:38:25 localhost dbus-daemon[2561]: /usr/lib64/python2.7/site-packages/gtk-2.0/gtk/__init__.py:57: GtkWarning: could not open display
Dec 28 08:38:25 localhost dbus-daemon[2561]: warnings.warn(str(e), _gtk.Warning)
Dec 28 08:38:25 localhost blueman-mechanism: Starting blueman-mechanism
Dec 28 08:38:25 localhost dbus[2561]: [system] Successfully activated service 'org.blueman.Mechanism'
Dec 28 08:38:25 localhost dbus-daemon[2561]: dbus[2561]: [system] Successfully activated service 'org.blueman.Mechanism'
Dec 28 08:38:27 localhost blueman-mechanism: reload 0 0
Dec 28 08:38:27 localhost kernel: Bridge firewalling registered
Dec 28 08:38:28 localhost dnsmasq[9879]: started, version 2.59 cachesize 150
Dec 28 08:38:28 localhost dnsmasq[9879]: compile time options: IPv6 GNU-getopt DBus no-i18n DHCP TFTP no-conntrack no-IDN
Dec 28 08:38:28 localhost dnsmasq-dhcp[9879]: DHCP, IP range 10.237.118.2 -- 10.237.118.254, lease time 1h
Dec 28 08:38:28 localhost dnsmasq[9879]: reading /etc/resolv.conf
Dec 28 08:38:28 localhost dnsmasq[9879]: using nameserver 192.168.0.1#53
Dec 28 08:38:28 localhost dnsmasq[9879]: read /etc/hosts - 3 addresses
Dec 28 08:38:29 localhost mgaapplet[9749]: running: ionice -p 9749 -n7
Dec 28 08:38:57 localhost blueman-mechanism: Exiting
Dec 28 08:38:57 localhost dbus-daemon[2561]: Starting blueman-mechanism
Dec 28 08:38:57 localhost dbus-daemon[2561]: reload 0 0
Dec 28 08:38:57 localhost dbus-daemon[2561]: Exiting
Dec 28 08:39:02 localhost CROND[12040]: (root) CMD ([ -d /var/lib/php ] && find /var/lib/php/ -type f -depth -mindepth 1 -maxdepth 1 -mmin +$(/usr/lib64/php/maxlifetime) -print0 | xargs -r -0 rm)
Dec 28 08:43:29 localhost mgaapplet[9749]: Computing new updates...
Dec 28 08:43:29 localhost mgaapplet[9749]: running: urpmi.update -a
Dec 28 08:44:04 localhost mgaapplet[9749]: updating inactive backport media Core Backports Testing (distrib9), Nonfree Backports Testing (distrib19), Tainted Backports Testing (distrib29), Core 32bit Backports (distrib37), Core 32bit Backports Testing (distrib39)
Dec 28 08:44:04 localhost mgaapplet[9749]: running: urpmi.update Core Backports Testing (distrib9)
Dec 28 08:44:05 localhost mgaapplet[9749]: running: urpmi.update Nonfree Backports Testing (distrib19)
Dec 28 08:44:07 localhost mgaapplet[9749]: running: urpmi.update Tainted Backports Testing (distrib29)
Dec 28 08:44:09 localhost mgaapplet[9749]: running: urpmi.update Core 32bit Backports (distrib37)
Dec 28 08:44:11 localhost mgaapplet[9749]: running: urpmi.update Core 32bit Backports Testing (distrib39)
Dec 28 08:44:29 localhost mgaapplet[9749]: Checking... Updates are available#012



and the mysql error log is below (couldn't get it to attach for some reason).

Code: Select all

111228  8:37:48 [Note] Plugin 'PBXT' is disabled.
111228  8:37:48 InnoDB: The InnoDB memory heap is disabled
111228  8:37:48 InnoDB: Mutexes and rw_locks use GCC atomic builtins
111228  8:37:48 InnoDB: Compressed tables use zlib 1.2.5
111228  8:37:48 InnoDB: Using Linux native AIO
111228  8:37:48 InnoDB: Initializing buffer pool, size = 80.0M
111228  8:37:48 InnoDB: Completed initialization of buffer pool
111228  8:37:48 InnoDB: highest supported file format is Barracuda.
111228  8:37:53  InnoDB: Waiting for the background threads to start
111228  8:37:54 Percona XtraDB (http://www.percona.com) 1.1.8-20.1 started; log sequence number 7469292468
111228  8:37:55 [Note] /usr/sbin/mysqld: ready for connections.
Version: '5.5.18-MariaDB'  socket: '/home/pnunn/.local/share/akonadi/socket-localhost.localdomain/mysql.socket'  port: 0  Mageia - MariaDB Community Edition (GPL)



Thanks

Peter.

looks all pretty clean.

but, every time you start it, it puts the mysql.err contents into mysql.err.old

so can you post your updated mysql.err.old file? you can add a .txt ext, iirc it should enable you to attach it.

after that, i think a new akonadi version might solve this issue (i donno if it's been submitted yet)

Crashed again... the mysql.err file had nothing much in it, other than normal operation stuff (in this file attached as well). The .old file is from the previous clash I guess.

Peter.

by any chance, did you remove the ib_log files? or did they dissappear on it's own?

contrary to what it looks like, it's not just a log file you can delete... it's a sort of log of what it still needs to do...

after the recovery, is everything still ok? can you except that it crashes (and likely akonadi restarted), see everything, or you can't read your email anymore?

No, i didn't remove any log files manually. The system seems to be working still, I can read the existing mail.

Peter.

do you mean that you can't get recent mail ?

No.. things are working fine, except for when I try and delete a message or close down kontact, then it crashes.

hum. it's really different to my original bug (which has been fixed with the akonadi 's upgrade).
Could you get a backtrace for this by attaching gdb to the kmail process & then delete a message ?

mikala wrote:Could you get a backtrace for this by attaching gdb to the kmail process & then delete a message ?

@pnunn: Maybe https://wiki.mageia.org/en/Debugging_software_crashes can help you in the process

ok in fact thoses bugs have been reported by others users :
https://bugs.kde.org/show_bug.cgi?id=289693 <-- the crash on exit
https://bugs.kde.org/show_bug.cgi?id=289967 <-- the crash when deleting an email.

Thanks guys,

does look as though it the same bug that's been reported.

The debugging tutorial is a great one too, thanks for that, I'll keep it handy.

For the time being I'll go back to thunderbird and see what transpires.

Have a great New Year.

Peter.

Guys,

you cannot ship Mageia with kontact the way it is.

Each time a new update comes out I try it again to see if its still broken or not.. it gets worse.

akregator does not show anything at all at the moment (although it works just fine stand alone).
kmail pops up a box "Unknown error. (Unable to fetch item for backend) each time I tried to do anything. It seems to be working (all be it VERY slowly) but the errors are consistent.

I know its not the packagers at fault, but does anyone have any contact with the developers? I can't believe how they have managed to screw up what was a good mail client so badly.

Peter.

i can't speak for mikala; but i've had several bugs with kmail on mga1, but the answer was always the same:
WONT_FIX

why? because upstream apparently obsoletes kmail1, while kmail2 was not usable according to mikala.

also kmail2 (what we will use now) is iirc quite the redo of the first one, so not the same at all...

i'm a longtime user of KDE and kontact, and it saddens me...

still, this is cauldron, we're shipping atm the beta version of KDE 4.8; so file as much bugs as you can upstream; so that they can get them fixed before KDE 4.8 final release...