On 08/05/2021 15:39, Carlos E. R. wrote:
On 08/05/2021 16.19, Sid Boyce wrote:
On 07/05/2021 22:28, Carlos E. R. wrote:
On 07/05/2021 22.58, Sid Boyce wrote:
I have a colleague (erstwhile) in Australia who ran into the problem and couldn't understand why.
One PC with a 40GB hard drive and another with a 40GB SSD both with the
same number of snapshots but only the SSD complaining of no space left.
After a lot of checking I asked him to delete a number of snapshots (reduced from 20 and 20 to 10 and 10) which now shows only 3.9GB used.
Later I'll file a bug as this could force someone who is not very Linux
savvy to do a fresh install or even move to another distro.
SSD or rotating rust has no difference. Further analysis would have found the real cause.
It is perfectly known issue that a small hard disk with snapshots is a
bad idea. You need at least 3 times the size you would normally use.
Understood. That's why I normally go for as large as I can get and over
time 1TB -> 2TB --> 5TB and I have a 6TB on standby in case of a failure.
The user installed openSUSE Leap because he doesn't like Windows or Ubuntu or other distros, added to that is the performance of modest hardware.
Yep.
The doc says >16GB and snapshots will be enabled which gives the impression that 40GB would hit no problems.
Sigh... My educated opinion is different...
What puzzled him was that the 40GB SSD experiencing the problem and the
40GB hard drive was OK .... obviously the SSD was loaded with more apps, videos, etc. than the hard drive.
I doubt videos would be a problem, they typically go to the /home partition. Unless there is no /home partition and is a directory, a practice I strongly dislike for "actual" use by people.
It could be more apps, or simply different zypper dup usage pattern.
If the total hard disk space is that limited (40GB) I would have selected a single ext4 partition, plus possibly swap. And Leap, unless having a specific reason for using factory. In that case, I would have ready the XFCE rescue USB.
Life was sweet until suddenly "zypper dup" complained.
Yes, you are not the first one hit by this issue. And using factory makes the issue worse as there are many more "dups".
I gather he accepted the installation defaults. I think he got started with Leap 14 and upgraded through to 15.2. After deleting half the snapshots this is what it now reports, used went from 41GB to only 3.9GB. linux-sl6n:~ # df -h Filesystem     Size Used Avail Use% Mounted on devtmpfs       3.9G 4.0K 3.9G  1% /dev tmpfs          3.9G     0 3.9G  0% /dev/shm tmpfs          3.9G 1.6M 3.9G  1% /run tmpfs          3.9G      0 3.9G  0% /sys/fs/cgroup /dev/sda2       41G 9.3G  31G 24% / /dev/sda2       41G 9.3G  31G 24% /var/opt /dev/sda2       41G 9.3G  31G 24% /var/lib/mariadb /dev/sda2       41G 9.3G  31G 24% /var/lib/libvirt/images /dev/sda2       41G 9.3G  31G 24% /var/lib/machines /dev/sda2       41G 9.3G  31G 24% /usr/local /dev/sda2       41G 9.3G  31G 24% /boot/grub2/x86_64-efi /dev/sda2       41G 9.3G  31G 24% /var/crash /dev/sda2       41G 9.3G  31G 24% /var/lib/mailman /dev/sda2       41G 9.3G  31G 24% /opt /dev/sda2       41G 9.3G  31G 24% /srv /dev/sda2       41G 9.3G  31G 24% /tmp /dev/sda2       41G 9.3G  31G 24% /var/spool /dev/sda2       41G 9.3G  31G 24% /var/lib/pgsql /dev/sda2       41G 9.3G  31G 24% /boot/grub2/i386-pc /dev/sda2       41G 9.3G  31G 24% /var/log /dev/sda2       41G 9.3G  31G 24% /var/lib/named /dev/sda2       41G 9.3G  31G 24% /.snapshots /dev/sda2       41G 9.3G  31G 24% /var/lib/mysql /dev/sda2       41G 9.3G  31G 24% /var/tmp /dev/sda4      424G  48G 377G 12% /home tmpfs          786M  12K 786M  1% /run/user/1000 /dev/sdb1       30G  22G 8.3G 73% /run/media/sav/Lexar Regards Sid. -- Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support Senior Staff Specialist, Cricket Coach Microsoft Windows Free Zone - Linux used for all Computing Tasks