'/proc/kcore' file is (apparently) 128T

Posted June 14, 2020 3.5k views

Hi, since a couple of weeks I have an unstable Wordpress site on a DigitalOcean droplet. The actual WP backup file is about 150MB. After investigating file sizes (df -h) it turns out that ’/dev/vda1/’ has used 20G (of 25G) and ’/proc/kcore’ is even 128T…

Is there a way to ‘shrink’ these files of delete old logs?


1 comment
  • Like all other files below /proc the kcore file is only a virtual file. It contains the RAM the kernel can allocate. Therefore this should not be touched or read. It is nothing to worry about. This file doesn’t use actual disk space and only exists virtually.

    On 64-bit systems the size of /proc/kcore is even 128TB because that’s the absolute limit of what 64-bit systems can allocate.

These answers are provided by our Community. If you find them useful, show some love by clicking the heart. If you run into issues leave a comment, or add your own answer to help others.

Submit an Answer
1 answer

Hi @olivervogt,

I’ll recommend finding out which service is using this process and restarting it. The process itself can’t hold 128T of space due to the fact the droplet doesn’t have as much.

As for the space issue, what I’ll recommend is using something in the midst of

This command will find the 5 biggest directories in your /

du -sh / | sort -n -r | head -n 5 

To display the largest folders/files including the sub-directories, run:

du -Sh / | sort -rh | head -5

If you want to display the biggest file sizes only, then run the following command:

find -type f -exec du -Sh {} + | sort -rh | head -n 5

I’ll recommend using these commands to try and find out where the biggest files are located and if they are logs truncate them or if anything else deal with them accordingly.

It’s possible your website has grown in space and you just need to upgrade our space. Having said that, you’ll need to decide that on the spot once you know where the space issues are coming from.


  • Thank you very much for the guidance - journalctl --vacuum-size=1G did reduce the filesize to an acceptable limit

    /proc/kcore is still huge… but the VM seems to be stable now.