Linux Disk Quota Exceeded for User – 3 Fixes

You're getting "disk quota exceeded" errors. I've seen this on RHEL 7, Ubuntu 22.04, and CentOS 8 — it's almost always a user filling up their allocated space. Here's the fix chain, from quickest to most thorough. Stop when it works.

30-Second Fix: Clear Trash and Temp Files

This gets you out of a jam 8 out of 10 times. Your home directory has a trash folder and temp cache that eat quota fast.

rm -rf ~/.local/share/Trash/* ~/.cache/* ~/.thumbnails/*
find ~/ -maxdepth 1 -name "*.core" -delete
find ~/ -maxdepth 1 -name "*.log" -delete

If you're using a shared server (like a school or corporate box), also check /tmp for files you own:

find /tmp -user $(whoami) -delete

When this works: You accidentally left a browser cache or download that filled your quota. You'll know it worked if you can write a file again.

Skip messing with ~/.bash_history — it's tiny and won't matter.

5-Minute Fix: Check Quota Limits and Identify the Hog

If the quick cleanup didn't cut it, you need to see your actual limits and what's eating space.

Check your quota

quota -s

Output looks like:

Disk quotas for user jdoe (uid 1001): 
Filesystem  blocks   quota   limit   grace   files   quota   limit   grace
/dev/sda1   45000*  50000   55000   7days   1200    2000    2500

The asterisk after 45000* means you're above the soft limit (50000 blocks). You have 7 grace days to get under it. The limit column (55000) is the hard cap — you can't exceed that at all.

Find the real space hog

du -sh ~/* ~/.[!.]* | sort -hr | head -10

This lists the 10 biggest items in your home dir, including hidden files. Common culprits:

~/.local/share/Steam — game installs can be 20+ GB
~/.npm — node_modules caches (I've seen 8GB)
~/snap — Snap packages on Ubuntu bloat fast
~/Downloads — old ISOs, tarballs
~/.docker — images and containers

Delete whatever you don't need with rm -rf. Be careful — double-check before blowing away ~/.ssh or ~/.config.

15+ Minute Fix: Move Files to Another Partition or Request a Quota Increase

You've cleaned everything, and you're still hitting the limit. Now you need a structural change.

Option A: Move data to a partition without quotas

Check if your system has a /scratch or /data mount that's not quota-limited:

df -h | grep -v quota
mount | grep -v quota

If you find one (like /scratch), move your big files there:

mv ~/big_project /scratch/$(whoami)/
ln -s /scratch/$(whoami)/big_project ~/big_project

Symlink saves you from changing paths in scripts.

Option B: Request a quota increase from the admin

Run repquota -a (only root can use it, but you can ask the admin to show you current usage). Then file a ticket like this:

User jdoe on server devbox1 hitting quota on /home (50000 blocks soft, 55000 hard). Current usage 48000 blocks, main hog is ~/research_data (15GB). Need soft limit bumped to 80000 for next 2 months. Can provide justification on request.

Admins appreciate specifics — don't just say "my quota's full."

Option C: If you have root (local machine), change the quota yourself

sudo edquota -u $(whoami)

This opens a vi screen. Edit the block soft and hard limits (in 1K blocks). For example, to give yourself 100MB soft and 120MB hard:

/dev/sda1  50000 102400 122880 ...

Save and exit. Then apply:

sudo quotaon -a

Check it:

quota -s

Warning: Setting quotas too high can crash the filesystem if disk fills up. Stay under 90% of total disk.

What Not to Waste Time On

Rebooting the machine — quotas are enforced at the filesystem level, not at boot. Wastes 5 minutes.
Running fsck — not a quota issue. Stops you from fixing the real problem.
Reinstalling software — unless you're tracking dirty data, it won't help.

Preventing This Again

Set a cron job to warn you at 80%:

# crontab -e
0 8 * * * /usr/bin/quota -s | grep -q "\*" && echo "Quota almost full!" | mail -s "Quota Warning" you@example.com

Or use ncdu (install it) to visually scan your home once a month:

ncdu ~/

That's it. You're back to work.