Cleaning Drives

This document should help you clean up drives that are running low on space.

Windows

• Tree Size Free can be used to analyze disk usage, allowing users to easily find locations that are using the most storage. Note that the free version shouldn't be used on Windows Server.

WSL

These instructions were adapted from the second-top response in this StackOverflow post

The entirety of storage used by a WSL distro appears as a single vdx drive in Tree Size and special steps need to be taken to shrink the volume.

1. Clean up contents of the WSL Distro using the steps in the following Linux section.

2. Stop Docker on your Windows machine, if it's running.

3. Stop WSL instances with the command wsl --shutdown

4. Create a file called compact-disk.txt with the following contents, making sure you reference the proper path instead of "CanonicalGroupLimited.UbuntuXX.NNNNNN":

   select vdisk file="C:\Users\%username%\AppData\Local\Packages\CanonicalGroupLimited.UbuntuXX.NNNNNN\LocalState\ext4.vhdx"
   attach vdisk readonly
   compact vdisk
   detach vdisk

5. Open an Admin Powershell instance

6. Change directory to the previously created file

7. Run the command diskpart /s compact-disk.txt

The final command will take a while to complete; up to an hour for very large drives.

Linux

du -ha | grep -P '^\S+G\s'

The du command can take a long time to run, so it's best to break up your disk usage analysis into smaller bites. For the most part, you will only need to check for disk usage in users' home directories, but there might be edge cases where you should look in other locations as well.

You can use the command du -ha | grep -P '^\S+G\s' to list disk usage in the current directory and filter out everything that isn't at least 1 GB in size (that's the grep modifier).

The output might look like this:

2.9G    ./.virtualenvs
1.5G    ./git/projects/aptim/chama22/chama_django/database/db.sqlite3
1.5G    ./git/projects/aptim/chama22/chama_django/database
1.5G    ./git/projects/aptim/chama22/chama_django
2.4G    ./git/projects/aptim/chama22
3.4G    ./git/projects/aptim
1.1G    ./git/projects/epa/scinote-web/node_modules
1.5G    ./git/projects/epa/scinote-web
1.1G    ./git/projects/epa/old_scinote
1.8G    ./git/projects/epa/ordscinote
5.4G    ./git/projects/epa
1.1G    ./git/projects/maiawater
12G     ./git/projects
1.2G    ./git/documentation/projects
1.5G    ./git/documentation
1.7G    ./git/gqc/shiny-react-testbeds/node_modules
1.9G    ./git/gqc/shiny-react-testbeds
1.4G    ./git/gqc/project-docs
1.2G    ./git/gqc/deepvibe-super-repo
7.9G    ./git/gqc
21G     ./git
28G     .

Common directories to consider purging might be:

1. Old or unused git repositories
2. Old or unused Python virtual environments
3. .cache/ directory content
4. Downloads/ directory content