For about a month nowI got that message « low space in home folder.
Cleaned my folder. It came back, and now the system cannot boot due to lack of space !!!
Ran ncdu on /and noticed there was something with var, ran ncdu on /var got the following :
4.6 GiB ##########] /log
523.3 MiB # ] /adm
494.7 MiB # ] /lib
376.1 MiB ] /cache
85.6 MiB ] /spool
30.7 MiB ] /tmp
16.0 KiB ] /run.runmove~
4.0 KiB ] /yp
e 4.0 KiB ] /opt
e 4.0 KiB ] /lock.lockmove~
e 4.0 KiB ] /gdm
4.0 KiB ] /games
e 4.0 KiB ] /crash
4.0 KiB ] .updated
@ 0.0 B ] lock
@ 0.0 B ] agentx
@ 0.0 B ] mail
@ 0.0 B ] run
Does the same on /log:here is what I got (part of it :
/…
2.2 GiB ##########] messages
2.2 GiB ######### ] warn
22.8 MiB ] /updateTestcase-2020-01-20-17-05-59
22.7 MiB ] /updateTestcase-2020-01-17-11-45-46
12.0 MiB ] /YaST2
10.0 MiB ] pk_backend_zypp-1
5.7 MiB ] boot.log
5.6 MiB ] zypper.log
4.4 MiB ] snapper.log
3.9 MiB ] /zypp
What am I to do ?
I’m on gnome, leap 15.1, kernel 4.12
I went into files and folders and found out :
there are two kernels : 4.15 and 4.16 that were downloaded when I used packagekit for updates. Need to delete these . But in what folder are there ???
Though I work with gnome, kde*plasma was installed years ago as I needed astronomy programs (kstar, stellarium) … how to softly uninstall that ?
Years ago, as I needed a program , 32 bits files were installed, though the whole system is in 64 bits. How to delete these withut disturbing the whole system ?
If this system is not an upgrade and installed BTRFS,
You are likely running out of disk space due to an accumulation of snapshots…
Open snapper and remove images you don’t think you’ll need.
Starting in 15.1,
A big change was made to the default layout,
Instead of separate partitions for root and /home,
it’s now all on one big partition (but separate volumes) so
Although snapshots are only of parts in /, your whole system’s disk usage is affected.
We need to know how much free disk space is present:
By simply inspecting /etc/fstab you should be able to determine if, the system partition is a Btrfs partition, or not.
If Btrfs then, with the user “root
”, “btrfs filesystem df /” will indicate if there’s a free space issue with the root partition, or not. - If not Btrfs then “df -h /” will indicate the root partition’s free space.
Looking on my system, I see that “/var/log” is only using 200M.
If I look just at "messages*, there is a file “/var/log/messages” and there are bunch of older files with names such as “messages-20200124.xz”. Those files are compressed version of earlier “messages” version.
It is usually safe to remove “/var/log/*.xz”. That deletes older log files. Unless you need to examine them for problems, you aren’t likely to need them. That could free up some space fairly quickly, unless those files are also in “btrfs” snapshots. Removing some unneeded older snapshots can also clear up space.
you are right. They are 2 folders with a size of 2.2 Gib each as you can see in firts post. I keep on deleting them but to no avail, each time they come back !
It seems that, the user “papaya” has an issue with directory usage – 279G space in their Home directory used.
Either, with that user or, the user “root
”, please execute the following CLI command: “find /home/papaya -size +200M -exec /usr/bin/ls -h -s {} ;” - If nothing is found, reduce the “size” parameter to 100M or, 50M or, even less, until the culprit or culprits, are discovered …
For the system directories, please use the user “root” to execute the CLI commands – the system privileges are needed to find the files which are using the partition’s space.
For the case of, “/var/log/”, the culprit could well be the systemd Journal.
Please execute, from the user “root”: “journalctl --disk-usage”.
If the systemd Journal is consuming more than a few GB, please consider vacuuming the Journal – assuming that, you do not need the Journal entries for administration purposes …
[INDENT=2]If you need older Journal entries, you’ll have to archive them before performing the housekeeping operation on the Journal.[/INDENT]
The command to reduce the partition space used by the systemd Journal is: “journalctl --vacuum-size=100M” «keeps 100M of Journal» or, “journalctl --vacuum-time=1weeks” «keeps 1 weeks worth of Journal» or, “journalctl --vacuum-files=5” «keeps only 5 Journal files».
Since openSUSE moved to “systemd” and the systemd Journal, ‘messages’ and ‘warn’ no longer exist in “/var/log/” – the Journal exists in “/var/log/journal/” …
Have you disabled systemd?
Are you still using SysV init?
AFAICS, the “sysvinit” RPM package is only available as source code …
You should be able to simply remove the “kstars” and “stellarium” either via YaST or “zypper remove” – the user “root” is needed to perform these tasks …
If you, use the YaST Software Manager to remove the RPM packages of the unwanted 32-bit applications, you’ll be prompted if there are dependencies to other installed RPM packages – you can then review those dependencies and make decisions about removing the 32-bit versions of the applications or, not …
The CLI command – also works with a “normal” user – “zypper search --installed-only -32bit” will display all the 32-bit packages installed on your system.
On this “clean” Leap 15.1 system, I currently have eighty five 32-bit packages installed – due to package dependencies – not because I need them …
A new default installation of Leap or Tumbleweed will produce both “messages and “warn” logs. They should be managed by “logrotate”, but its configuration can become confused during upgrades. However those large files can be safely removed, although I would be surprised if the were the primary cause of the full partition.
You should be able to simply remove the “kstars” and “stellarium” either via YaST or “zypper remove” – the user “root” is needed to perform these tasks …
There will still be a lot of unwanted KDE/Plasma stuff. YaST can also be used to safely, but painfully, remove the orphans.
It seems that, the user “papaya” has an issue with directory usage – 279G space in their Home directory used.
Either, with that user or, the user “root
”, please execute the following CLI command: “find /home/papaya -size +200M -exec /usr/bin/ls -h -s {} ;” - If nothing is found, reduce the “size” parameter to 100M or, 50M or, even less, until the culprit or culprits, are discovered …
For the system directories, please use the user “root” to execute the CLI commands – the system privileges are needed to find the files which are using the partition’s space.
For the case of, “/var/log/”, the culprit could well be the systemd Journal.
Please execute, from the user “root”: “journalctl --disk-usage”.
If the systemd Journal is consuming more than a few GB, please consider vacuuming the Journal – assuming that, you do not need the Journal entries for administration purposes …
[INDENT=2]If you need older Journal entries, you’ll have to archive them before performing the housekeeping operation on the Journal.[/INDENT]
The command to reduce the partition space used by the systemd Journal is: “journalctl --vacuum-size=100M” «keeps 100M of Journal» or, “journalctl --vacuum-time=1weeks” «keeps 1 weeks worth of Journal» or, “journalctl --vacuum-files=5” «keeps only 5 Journal files».
journalctl --disk-usage
Archived and active journals take up 446.0M in the file system
I am confused. Your title is "No space on /home!!!”, yet you show some investigations on “/var” and nothing on “/home”. It seems that the root partition is full, but you omit the usage off “/”.
I would like to see the output of
> df -hT |grep -v tmpfs
and either
# ncdu -x /
# ncdu -x /home
or
# du -shc /* 2>/dev/null
# du -shc /home/* 2>/dev/null
If your system is not cleaning up /var/log then perhaps it is also not cleaning up /tmp.