[SOLVED] after dist-upgrade, system does not boot.

kaz2100 · 10-09-2014, 12:32 AM

Hya,

Probably I am hitting one of the problems hanging around in Debian world.

System
Debian jessie, AMD64, Core2 duo, (Kernal 3.14.12 custom)

What happened
After dist-upgrade, system did not boot, (as the title reads) and ran into emergency mode.

Last system upgrade was 6/Aug/2014

Troubleshooting
Bootup message suggests that some trouble with mounting disks.
journalctl -xb reads "timed out ... uuid ..."
However, disk is alive under emergency mode!

It turned out to be the trouble is non-root disk (SSD mounted as ext2), related fstab entry looks like

Code:

UUID=1604cd4a-9f01-41a6-bbc5-.........      /path/to/mount/point       ext2    errors=remount-ro       0       1

Once the line is commented out, system boots, but still following message on bootup.

Code:

Dependency failed for /dev/disk/by-uuid/....

I am still working on this issue. Maybe this thread is same nature.

EDDY1 · 10-09-2014, 01:28 AM

Have you checked blkid to see if entry is correct?

kaz2100 · 10-09-2014, 09:20 PM

Hya

Thanks for your reply.

It looks like that blkid is OK.

I am not sure if this thread is rlated. An option "auto" discussed.

By the way, there is another problem. This jessie is a comute node of a cluster (by slurm). Currently it is off for some reason.

The package slurm-llnl also chokes on upgrade. This is where it chokes.

Code:

.
.
+++ case "$FANCYTTY" in
+++ true
+++ echo -n '[....] '
[....] +++ '[' -z slurm-llnl.service ']'
+++ echo -n 'Starting slurm-llnl (via systemctl): slurm-llnl.service'
Starting slurm-llnl (via systemctl): slurm-llnl.service+++ log_daemon_msg_post 'Starting slurm-llnl (via systemctl)' slurm-llnl.service
+++ :
+++ /bin/systemctl start slurm-llnl.service
Warning: Unit file of slurm-llnl.service changed on disk, 'systemctl daemon-reload' recommended.

I somewhat think everything is systemd related.

cheers

EDDY1 · 10-09-2014, 11:55 PM

Did you install systemd-sysv

kaz2100 · 10-10-2014, 03:22 AM

Hya

Thanks for your comment.

Yes, systemd-sysv is installed.

For some reason, sysvinit is installed also, and deborphan says there is no reason for sysvinit to be there. I removed it, but the original problem persists.

Another wierd thing, I noticed is

Code:

# aptitude show systemd-sysv
Package: systemd-sysv                    
New: yes
State: installed
Automatically installed: yes
Version: 215-5+b1
Priority: important
Section: admin
Maintainer: Debian systemd Maintainers <pkg-systemd-maintainers@lists.alioth.debian.org>
Architecture: amd64
Uncompressed Size: 73.7 k
Depends: systemd (= 215-5+b1)
PreDepends: systemd
Conflicts: sysvinit-core, sysvinit-core, upstart, upstart, systemd-sysv
Replaces: sysvinit (< 2.88dsf-44~), sysvinit (< 2.88dsf-44~), sysvinit-core,
          sysvinit-core, upstart, upstart
Description: system and service manager - SysV links
 
Homepage: http://www.freedesktop.org/wiki/Software/systemd

Why do packages (conflicting and replacing) appear twice?

cheers

EDDY1 · 10-10-2014, 03:28 AM

Have you tried updating to see if the updates were complete?

kaz2100 · 10-10-2014, 03:49 AM

Hya,

Thanks for your reply.

Yes, I have done that several times.

cheers

EDDY1 · 10-10-2014, 04:04 AM

What was the mountpoint for the ssd?

kaz2100 · 10-13-2014, 07:45 PM

Hya,

Thanks for your reply. Apology for delay.

It used to be /dev/sdb1. It is also sdb1 according to /etc/blkid.tab.

cheers

jlinkels · 10-13-2014, 08:45 PM

Wild shot...
I had this problem (/dev/disk-by-uuid...) this weekend on Debian Unstable. The machine did boot in the old 3.2.0 kernel. I reinstalled the current kernel (linux-image-amd64 I believe) and then the error was gone.

On my machine it could have been caused because my son had been messing around with nvidia drivers.

jlinkels

kaz2100 · 10-13-2014, 09:13 PM

Hya

Thanks for your post.

I have up-to-date kernel (3.16.3). Also, dist-upgrad is up-to-date.

EDDY1 · 10-13-2014, 09:39 PM

My only problem with systemd &mount points was having /usr on separate partition

jlinkels · 10-14-2014, 06:50 AM

Quote:

Originally Posted by kaz2100

Hya

Thanks for your post.

I have up-to-date kernel (3.16.3). Also, dist-upgrad is up-to-date.

Me too. I said I reinstalled the kernel despite of everything being up-to-date.

Code:

apt-get install --reinstall linux-image-amd64

jlinkels

kaz2100 · 10-14-2014, 09:13 PM

Hya,

Correction!

Its mount point is /home/pathToMount, not /dev/sdb1.

Apology!

cheers

kaz2100 · 10-26-2014, 05:44 PM

Hya

Self reply.

I guess most probably this issue is related to this or this.

I will investigate.

cheers