Slow startup of Plasma 6.8 (again)

snack · Today 07:37:39

After the upgrade to Plasma 6.8 I have again issues with very slow startup time after login, as already happened with 6.4 and described here. But this time I disabled all the entries in the fstab, leaving only three like this one:

192.168.178.52:/home/nicola                        /mnt/conan/home nfs             rw,user,noatime,noauto 0 0

which however having the `noauto` flag should not interfere. So I guess it's probably something different and not related to disk mount. Did someone else experience the same after upgrading to 6.8?

stefluik · Today 11:24:01

yes I feel the same issue.

seth · Today 13:55:52

Well, a #meetoo thread won't get to the bottom of this, https://wiki.archlinux.org/title/Genera … mpty_posts

=> Please post your complete system journal for the boot:

sudo journalctl -b | nc termbin.com 9999

so we can see where and why things stall.

snack · Today 14:06:50

Thanks seth, I think tehre's some problem with termbin:
```
$ sudo journalctl -b | nc termbin.com 9999
https://termbin.com/ru859
read(net): Connection reset by peer
```

Edit: the hello world example on termbin.com works, so I guess there might be some character in the output of journalctl that screws the connection (??), or maybe I am exceeding a maximum limit of paste-able characters?

Last edited by snack (Today 14:17:34)

seth · Today 14:25:22

Journal is 8400 lines, you probably got quota'd

ott 22 09:10:36 stryke systemd-logind[502]: New session '1' of user 'sddm' with class 'manager-early' and type 'unspecified'.
ott 22 09:10:43 stryke sddm[537]: Session "/usr/share/wayland-sessions/plasma.desktop" selected, command: "/usr/lib/plasma-dbus-run-session-if-needed /usr/bin/startplasma-wayland" for VT 1
ott 22 09:10:43 stryke systemd-logind[502]: New session '2' of user 'mori' with class 'user' and type 'wayland'.
ott 22 09:10:43 stryke systemd-logind[502]: New session '3' of user 'mori' with class 'manager' and type 'unspecified'.
…
ott 22 09:10:44 stryke kwin_wayland_wrapper[1386]: Detected locale "LC_CTYPE=C;LC_NUMERIC=en_US.UTF-8;LC_TIME=it_IT.UTF-8;LC_COLLATE=C;LC_MONETARY=it_IT.UTF-8;LC_MESSAGES=C;LC_PAPER=C;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=it_IT.UTF-8;LC_IDENTIFICATION=C" with character encoding "ANSI_X3.4-1968", which is not UTF-8.
…
ott 22 09:10:48 stryke systemd[1]: Startup finished in 16.169s (firmware) + 2.656s (loader) + 3.231s (kernel) + 1.506s (initrd) + 17.347s (userspace) = 40.911s.
……
ott 22 09:10:50 stryke dhcpcd[1304]: eth0: leased 172.16.242.76 for 43200 seconds
ott 22 09:10:50 stryke dhcpcd[1304]: eth0: adding route to 172.16.0.0/16
ott 22 09:10:50 stryke dhcpcd[1304]: eth0: adding default route via 172.16.0.254
ott 22 09:10:50 stryke systemd[1]: Stopping Automatic wireless network connection using netctl profiles...
ott 22 09:10:50 stryke kernel: wlan0: deauthenticating from d8:24:bd:2e:ab:00 by local choice (Reason: 3=DEAUTH_LEAVING)
ott 22 09:10:51 stryke dhcpcd[1336]: wlan0: carrier lost
ott 22 09:10:51 stryke dhcpcd[1336]: received SIGTERM, stopping
ott 22 09:10:51 stryke dhcpcd[1336]: wlan0: removing interface
ott 22 09:10:51 stryke dhcpcd[1336]: received SIGTERM, stopping
ott 22 09:10:51 stryke dhcpcd[1336]: dhcpcd exited
ott 22 09:10:51 stryke systemd[1]: netctl-auto@wlan0.service: Deactivated successfully.

SDDM started 09:10:36, 09:10:43 you're logging in, 09:10:44 kwin starts, 09:10:48 systemd thinks the startup has finished (most time in userspace, I suspect it's because of the wifi/eth jugglings?)

very slow startup time after login

Can you please describe the symptoms of that and quantify "very slow" in seconds?

I suspect the cause is around

ott 22 09:10:33 stryke systemd[1]: Unnecessary job was removed for /sys/subsystem/net/devices/eth0.
ott 22 09:10:33 stryke systemd[1]: netctl-ifplugd@eth0.service: Bound to unit sys-subsystem-net-devices-eth0.device, but unit isn't active.
ott 22 09:10:33 stryke systemd[1]: netctl-ifplugd@eth0.service: Job netctl-ifplugd@eth0.service/start failed with result 'dependency'.
ott 22 09:10:33 stryke (udev-worker)[372]: eth0: Process '/usr/bin/systemctl restart netctl-ifplugd@eth0.service' failed with exit code 1.
…
ott 22 09:10:37 stryke kernel: r8152 4-1.4:1.0 eth0: carrier on

the ethernet dongle?

What if you disable the wifi or plug the ethernet dongle early or disable all netctl services?

snack · Today 14:30:52

The symptoms are that after entering username and password in the login screen the black welcome screen with the rotating gear stands there for about 15 seconds vs 2-3 seconds with Plasma 6.7.
This happens also with the ethernet dongle already plugged on power on (I'm not sure if this goes in the direction of your "plug the ethernet dongle early" suggestion). I'll disable all the network units and check on next reboot, right now I'm at work and cannot reboot.

Thanks for your help!

seth · Today 14:44:03

So this here

ott 22 09:10:54 stryke systemd[1]: user-975.slice: Consumed 1.746s CPU time over 17.574s wall clock time, 567.4M memory peak.
ott 22 09:10:57 stryke systemd[1]: dbus-:1.1-org.kde.powerdevil.discretegpuhelper@0.service: Deactivated successfully.
ott 22 09:10:57 stryke systemd[1]: dbus-:1.1-org.kde.powerdevil.chargethresholdhelper@0.service: Deactivated successfully.
ott 22 09:10:57 stryke systemd[1]: dbus-:1.1-org.kde.powerdevil.backlighthelper@0.service: Deactivated successfully.
ott 22 09:11:13 stryke kdeconnectd[1847]: Could not query capabilities from notifications server
ott 22 09:11:14 stryke systemd[1]: systemd-localed.service: Deactivated successfully.
ott 22 09:11:17 stryke systemd[1346]: Starting User preferences database...
…
ott 22 09:12:17 stryke systemd[1346]: dbus-:1.2-org.kde.KSplash@1.service: Main process exited, code=exited, status=1/FAILURE
ott 22 09:12:17 stryke systemd[1346]: dbus-:1.2-org.kde.KSplash@1.service: Failed with result 'exit-code'.

There's a metric shit-ton of

ott 22 09:11:53 stryke systemd[1346]: Started Launch DrKonqi for a systemd-coredump crash (PID 1515/UID 10085).
ott 22 09:11:53 stryke drkonqi-coredump-launcher[3838]: Detected locale "LC_CTYPE=C;LC_NUMERIC=en_US.UTF-8;LC_TIME=it_IT.UTF-8;LC_COLLATE=C;LC_MONETARY=it_IT.UTF-8;LC_MESSAGES=C;LC_PAPER=C;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=it_IT.UTF-8;LC_IDENTIFICATION=C" with character encoding "ANSI_X3.4-1968", which is not UTF-8.
                                                        Qt depends on a UTF-8 locale, and has switched to "C.UTF-8" instead.
                                                        If this causes problems, reconfigure your locale. See the locale(1) manual
                                                        for more information.
ott 22 09:11:53 stryke drkonqi-coredump-launcher[3838]: Unable to find file for pid 154938 expected at "kcrash-metadata/baloo_file_extractor.407189d739d049328f2a50e454bd9175.154938.ini"

so some stale backtraces but that's well after

ott 22 09:11:17 stryke Discord[1681]: App ready

(which makes up for the rest of the journal)

=> try to disable kdeconnectd ?

I can't see where the notification server (plasmashell) would fail nor other dbus errors, but there's https://bbs.archlinux.org/viewtopic.php?id=315215 which happens because baloo starts before/without the wayland environment set.
I cannot really imagine how this could pertain to the session bus unless you've introduced some dbus-launch somewhere?

printenv

Possibly unrelated, but replace pulseaudio w/ pipewire-pulse

Arch Linux

#1 Today 07:37:39

Slow startup of Plasma 6.8 (again)

#2 Today 11:24:01

Re: Slow startup of Plasma 6.8 (again)

#3 Today 13:55:52

Re: Slow startup of Plasma 6.8 (again)

#4 Today 14:06:50

Re: Slow startup of Plasma 6.8 (again)

#5 Today 14:25:22

Re: Slow startup of Plasma 6.8 (again)

#6 Today 14:30:52

Re: Slow startup of Plasma 6.8 (again)

#7 Today 14:44:03

Re: Slow startup of Plasma 6.8 (again)

Board footer