I just installed updates from the Leap Current repo on an external 60 GB hard drive. There was over 1900 updates and towards the end, zypper was telling me there wasn’t enough space to install the last few packages. I began investigating and found that /home (31GB) was only 9% full while / (21GB) was 100% full. According to Dolphin, /proc is 128 TB. That’s right, terabytes. Both of the directories /dev and /run say there is 3.8 GB of free space. Even after deleting an older kernel, I still didn’t reclaim any space.
Does anyone know what’s going on here? Obviously I can’t be using so much space with just a basic install of Leap plus updates.
Also, the YaST partitioner tool says it can’t shrink or expand BTRFS partitions. Is this true?
Well /proc is not a real directory on disk is is virtual, it is the systems process. All thing in Unix and Linux are files. In addition /dev and /sys are only in existence on a running system it does not exist on the drive.
As to space I assume you installed the default file system which is BTRFS for root and this file system uses a program called snapper to take periodic snap shots of the file system. These snap shots do take up space so it is now the recommendation to have a minimum of 40 gigs for root or turn off snapper or at least cut down the amount of snaps it takes. 20 gig is not enough but still more then enough not to turn it off
You can shut it down from Yast but command line is better all around so do a man snapper for the manual instructions
On 2015-09-20 02:26, pilotgi wrote:
>
> I just installed updates from the Leap Current repo on an external 60 GB
> hard drive. There was over 1900 updates and towards the end, zypper was
> telling me there wasn’t enough space to install the last few packages. I
> began investigating and found that /home (31GB) was only 9% full while /
> (21GB) was 100% full.
Well, you need more space on “/” or free enough. Typical on btrfs.
> According to Dolphin, /proc is 128 TB. That’s
> right, terabytes.
Ignore it. Not a bug. It is not a real file, but a virtual one the size
of the entire possible memory space of your CPU.
> Both of the directories /dev and /run say there is 3.8
> GB of free space.
Ignore them. Virtual filesystems.
> Even after deleting an older kernel, I still didn’t
> reclaim any space.
>
> Does anyone know what’s going on here? Obviously I can’t be using so
> much space with just a basic install of Leap plus updates.
Yes, you can. With snapshots. I can’t explain how to clear them, but
there are plenty of threads here about it.
Did you originally install Leap Milestone 1 or 2? Guessing 1 and you’re trying to update to 2, and by what means? What is meant by “Leap Current repo”, please provide the link?
Please post (in code tags) the content of snapper’s root config file to be found in /etc/snapper/.
BTW 20GB should be more than enough for testing Leap, depending on what you have in that config file.
No it isn’t normal, and I have never had that problem even on Tumbleweed which gets a lot of updates regularly. So let us see your snapper config as requested.
I estimate you were very close to the Beta1 release, where for example my zypper dup from M2 to Beta1 accounted for about 2000 changes. If you go up to the parent(s) of that repo you will see the Beta1 repo. Personally, I wouldn’t have used 42.1-Current as it offers a volatile target for testing and comparison with others. Either way, given the massive update in terms of filling your root partition on Btrfs, you were possibly heading for trouble. It sounds like you nearly got away with it.
I managed it on a 20GB partition with a reduced snapper config, down from the default. I also need to be on my Leap partition to check a few things, and to compare mine with your actual config - to come later:
Edit: Just realised I’m not on the machine I was having the problem with. I’ll post the config file tonight.