Last du screwed my system - again

[ghstryder]

I'm guessing these are the relevant entries from journalctl -b -p err

Code Select Expand


ct 09 07:32:23 p690 systemd[3039]: Failed unmounting /sys/fs/fuse/connections.
Oct 09 07:32:23 p690 systemd[3039]: Failed unmounting /dev/mqueue.
Oct 09 07:32:23 p690 systemd[3039]: Failed unmounting /dev/hugepages.
Oct 09 07:32:23 p690 systemd[3039]: Failed unmounting /sys/kernel/debug.
Oct 09 07:32:23 p690 systemd[3039]: Failed unmounting /run/user/118.
Oct 09 07:32:23 p690 systemd[3039]: Failed unmounting /home.


[musca]

Hello,

now also the second issue got adressed:

  * Revert "core: add a "Requires=" dependency between units and the
    slices they are located in". This causes user systemd instances to try and
    unmount system mounts (and succeed if you login as root). (Closes: #801361)

systemd, systemd-sysv, libsystemd0, libnss-myhostname, libpam-systemd, udev, libudev1 have been uploaded in version 227-2 and packages will be available from the usual mirrors after the next sync (about 15:30 UTC, i think).

@ghstryder: yes.

greetings
musca

[ghstryder]

Thanks musca!

After a couple of minor changes, all is well.

162 upgraded, 0 newly installed, 0 to remove and 0 not upgraded

[terroreek]

Lightdm and Siduction are booting up correctly but I am still seeing that systemd-rfkill.service is timing out. 

-edit- looks like systemd-rfkill.service is tied to my bluetooth adapter. 

[clubex]

I upgraded systemd to 227.2 this morning and everything seems OK. My /home is now mounted and I can get to the desktop.

I'm  still getting the following errors:

Code Select Expand


Oct 10 09:46:32 westfield2 systemd[1]: rpcbind.service: Job rpcbind.service/start deleted to break ord$
Oct 10 09:46:34 westfield2 systemd[1]: rpcbind.service: Job rpcbind.service/start deleted to break ord$
Oct 10 09:46:40 westfield2 kernel: k10temp 0000:00:18.3: unreliable CPU thermal sensor; monitoring dis$
Oct 10 09:47:00 westfield2 colord-sane[1090]: io/hpmud/pp.c 627: unable to read device-id ret=-1


Ignoring the last two errors (the sensor one is well known and is as old as the moon and the colord-sane one is probably because the scanner isn't connectted) I'm still getting the rcpbind.sefvice errors even though:

Code Select Expand


systemctl status rcpbind.service
● rcpbind.service
   Loaded: not-found (Reason: No such file or directory)
   Active: inactive (dead)


The major breakdown seem to me to be fixed .

That said I think  we should consider whether systemd is so too deeply integrated into our systems. For me having a  systemd bug cause such a catastrophic breakdown and there be now way I can fix it except to wait for an update to systemd is unexceptable, Perhaps this a matter for opening another topic?

Anyway if no one objects I'll close this topic as solved tomorrow morning (11th Oct).

[devil]

You have to add in the favour of the two Debian Maintainers of systemd (Michael Bibl & Martin Pitt) that they react very fast. Maybe they do release new versions too fast as well. But that is Debian Unstable.


greetz
devil

[seasons]

I'm still getting the rcpbind.service errors

https://ask.fedoraproject.org/en/question/49946/how-to-delete-a-systemd-service/?answer=49954#post-id-49954