Finding 250GB of Missing Storage On My Mac: A Warning For Large Dataset Users

I recently faced a puzzling issue: my 1TB MacBook Pro showed only 150GB free, but disk analyzers could only account for about 500GB of used space. After hours of troubleshooting, I discovered that Spotlight’s search index had balooned to 233GB, hundreds of times larger than normal.

The Problem

Standard disk analyzers showed that my mac had 330GB of “Inaccessible Disk Space” and 66GB of “Purgeable Disk Space” but no clear explanation for where my storage went. Removing the purgeable space was easy enough with sudo purge but none of the recommended fixes from ChatGPT like clearing Time Machine snapshots, clearing unused conda packages with pip cache purge and conda clean --all, and restarting the computer had any effect on the inaccessible disk space.

The Discovery

Finally, after hours of researching online and searching all files with administator privilages I found the culprit: .Spotlight-V100 was using 233GB. A normal Spotlight index should be 1-10GB. The likely cause: I work with large datasets containing hundreds of thousands of nested folders, each with multiple subfolders and files. Spotlight attempts to index every single file and folder, and with this volume of items, the index had grown massively out of proportion. It wasn’t just the file contents but the sheer number of directory entries that caused the bloat.

The Solution

# Force Spotlight to rebuild
sudo mdutil -E /
# Disable Spotlight on all volumes
sudo mdutil -a -i off
# Re-enable Spotlight on all volumes
sudo mdutil -a -i on
# Force rebuild again
sudo mdutil -E /

The commands took a while to complete but eventually freed up the space

Prevention

To prevent this from happening again, you can exclude folders from Spotlight indexing. Apple provides instructions for this: Apple Support

If you work with large datasets containing hundreds of thousands of nested folders, it’s worth adding those directories to Spotlight’s exclusion list. If you’re missing significant disk space on macOS and work with large datasets, check your Spotlight index size, it might be the problem!

Author