I had a functioning Leap 15.1 system. I installed apcupsd and enabled it to start at boot. My UPS has a serial cable. For about a week, I got messages that communication with the UPS was lost within about a minute of the desktop’s loading. Now the system will not boot to the desktop and I get a message that the system cannot communicate with /dev/tty0. I am told to type the root password to do maintenance. If I do that and then type yast, yast starts, but not as root, so I can’t do anything to the system, I am also given the option to press Control-D to continue to boot, but that just sends me the same message after a couple of minutes.
I don’t care about the serial port or whether the system can communicate with the UPS. I can’t figure out how to start Yast as root so that I can set apcupsd to start manually.
A different question. My root partition is full. How can I delete some files from the /tmp directory so that the system has room to write error messages?
I’ve been using OpenSuSE for quite a while, but I’m not really a power user, obviously. Any help will be appreciated.
You may have to boot some kind of rescue media in order to mount / rw. If / can be booted rw, start with deleting files in /var/log/journal// if more than a small handful of recent files is present there. Anything more than a few boots old there is rather unlikely to be useful for anything but wasting space. Next look through /var/cache/zypp/ using ncdu. Old rpms are good space wasters that can be removed with ‘zypper clean <options>’. /var/log/.xz files are all but useless as well. Do you have more than two kernels in /boot/? They are huge space gobblers. I use mc as root user for most find and delete chores, while ncdu is better for evaluating deep directory nests.
Thanks, Felix. There were a lot of .xz files in /var/log, and I deleted them. I still get the same message when trying to boot to the default system, though. Maybe it’s time for a reinstall. Not one of my favorite things.
.xz files are more of a take up more space onscreen in /var/log/ directory listing problem than a significant disk space problem. Did you not find more than two kernels installed? As long as the currently booted one is working, none of the old ones are needed. Old kernels and their module trees are the biggest space wasters.
I tried creating another installation on a different hard drive, but found that I couldn’t read the files on the hard drive with the problem installation on it very well. Then I downloaded and burned a live DVD, which does allow me to see the files with Dolphin. However, I can’t find any more than two kernels in /boot. I’m assuming that a couple of files that are 8 mb in size are kernels. There is nothing in the /tmp directory and there is no /var/log directory that I can see. I really don’t know what’s taking up so much space. I also tried to search the partition for .rpm files, but could find none.
A lot of the guidance that I’ve seen for open SUSE installations says to make partitions of a few gb for /boot, /, and /var, but this 80gb partition that has those mount points is full. The next time I do an installation I’ll put most of the space on the drive into /, /boot/ and /var, and use a different drive for /home. This is a workstation with office files and some multimedia programs that didn’t seem to me to be such a space hog, but is.
I’m not terribly familiar with what linux saves and where it puts it, so I’m reluctant to just delete stuff. The learning curve is steeper than I imagined, though.
I can read fstab with vim and run those commands, but I can’t figure out how to direct the output to a USB stick, or save it in any way. Please suggest a way to write the output as a text file and copy it to usb stick /dev/sdb5.
(One of us should have suggested this before now.) Before you proceed with prior request, read https://en.opensuse.org/SDB:BTRFS and determine whether your primary filesystem is on BTRFS. If it is, you’ll want to determine if snapshots are the reason for the lack of space. If yes, remove the old ones and you should be back in business. While you’re at it, follow any other of its suggestions that it seems prudent to do at this time.
If EXT4 rather than BTRFS, then:
Boot 15.1 installation media
Choose from second screen Rescue System
Boot
Login root
# dmesg -w
Insert another stick in another USB port
Note when the kernel finds the stick its device name(s)
Ctrl-C to exit dmesg
# fdisk -l
to discover partitioning and to choose both / and stick filesystems to mount and write to
# mount /dev/sdXY /mnt # (mount EXT4 / filesystem)
# cd /tmp
# fdisk -l > fdiskout.txt
# blkid > blkidout.txt
# lsblk > lsblkout.txt
# cat /mnt/etc/fstab > fstabcat.txt
# umount /mnt
# mount /dev/sdS# /mnt # (mount a stick filesystem)
# cp -a *txt /mnt
[QUOTE=mrmazda;2931740](One of us should have suggested this before now.) Before you proceed with prior request, read https://en.opensuse.org/SDB:BTRFS and determine whether your primary filesystem is on BTRFS. If it is, you’ll want to determine if snapshots are the reason for the lack of space. If yes, remove the old ones and you should be back in business. While you’re at it, follow any other of its suggestions that it seems prudent to do at this time.
It was BTRFS. I didn’t know about .snapshots. I believe that the / partition was simply full of snapshots. Unfortunately, from an emergency boot, if you try to delete snapshot 1, the system will thereafter only boot to a grub menu. I seem to learn everything the hard way.