OrbStack Taking Too Much Disk Space On Mac?
OrbStack can look much larger than `docker system df` because VM storage, sparse allocation, build cache, images, and volumes are different buckets. The safe move is to measure first, then clean cache before touching volumes or resetting state.
No mail app or GitHub login? Send this directly from any inbox.
liuminsheng3@gmail.com - SafeDisk OrbStack Docker Cleanup Policy Invoice
OrbStack looks huge? Email before resetting VM state or pruning volumes.
Send the size mismatch you see between OrbStack and Docker. We will reply with the first safe read-only check before anyone touches local databases, volumes, or VM state.
curl -fsSL https://site-xi-orcin-50.vercel.app/scan.sh -o safedisk-scan.sh && bash safedisk-scan.sh "$HOME/Desktop"
No mail app or GitHub login? Send this directly from any inbox.
liuminsheng3@gmail.com - SafeDisk OrbStack Docker Cleanup Policy Invoice
Use this before anyone resets OrbStack.
Paste this into a team thread or support issue when people see a huge OrbStack folder but the Docker numbers do not match. It keeps the first action read-only and protects local volumes.
OrbStack disk usage can include VM storage, sparse allocation, images, build cache, and volumes. Please do not reset OrbStack or prune volumes yet.
First collect read-only evidence:
du -sh "$HOME/.orbstack" "$HOME/OrbStack" 2>/dev/null
find "$HOME/.orbstack" "$HOME/OrbStack" -maxdepth 4 -type f -size +1G -print0 2>/dev/null | xargs -0 du -h 2>/dev/null | sort -hr | head -30
docker system df -v
docker builder du 2>/dev/null | tail -60
Clean rebuildable build cache and unused images first. Review named volumes before deleting anything because they may contain local databases, fixtures, or project state.Send the OrbStack evidence and get the safest first cleanup step.
Paste the exact disk numbers you see, whether local databases or volumes matter, and what you already tried. We will reply with the lowest-risk cleanup order before reset or volume prune.
Want a team cleanup policy instead of one-off pruning?
Request the $99 team pilot invoice. We start from one representative Mac, identify safe/review/protected buckets, and turn it into a repeatable OrbStack/Docker cleanup policy within 48 hours.
Start With Read-Only Measurements
du -sh "$HOME/.orbstack" "$HOME/OrbStack" 2>/dev/null
find "$HOME/.orbstack" "$HOME/OrbStack" -maxdepth 4 -type f -size +1G -print0 2>/dev/null | xargs -0 du -h 2>/dev/null | sort -hr | head -30
docker system df -v
docker builder du 2>/dev/null | tail -60Usually Lower Risk
- Old build cache shown as reclaimable by Docker or BuildKit.
- Unused images that can be pulled or rebuilt again.
- Stopped containers from inactive projects.
Review First
- Docker volumes, especially local Postgres, MySQL, Redis, object stores, and fixtures.
- OrbStack VM data or disk images, because they can contain active container state.
- Resetting OrbStack, because it may recover space but can wipe local state.
Safer Cleanup Order
- Save `docker system df -v` output so volumes and reclaimable cache are visible.
- Clean build cache first if it is clearly old and rebuildable.
- Remove unused images after checking active projects.
- Back up or export database volumes before pruning volumes.
- Only consider reset after you know what state will be lost.
Get an OrbStack cleanup diagnosis
The SafeDisk scan now includes OrbStack, Docker Desktop, Rancher Desktop, Docker build cache, and common developer storage paths.
curl -O https://site-xi-orcin-50.vercel.app/scan.sh && bash scan.sh