Disk and cloud storage analyzer with caching, CLI, and web UI.
pip install disk-treeStart the server and open http://localhost:5001:
disk-tree-serverView all cached scans, start new scans for local paths or S3 buckets:
Browse directories with size, modification time, children, and descendant counts. Multi-select with keyboard navigation, bulk delete:
Interactive treemaps for visualizing space usage:
Browse and scan S3 buckets:
disk-tree index --help
# Usage: disk-tree index [OPTIONS] [URL]
#
# Options:
# -C, --no-cache-read Force fresh scan (ignore cache)
# -g, --gc Garbage collect old scans
# -s, --sudo Run gfind with sudo
# -m, --measure-memory Track peak memory usage
# --help Show this message and exit.
disk-tree scans # List cached scans (JSON)
disk-tree-server # Start the web UI serverScan an S3 bucket:
disk-tree index s3://ctbk
# 2333 files in 138 dirs, total size 45.1G
# 4B test.txt
# 66.1K favicon.ico
# 3.9M index.html
# 5.5M static
# 11.2M .dvc
# 95.0M tmp
# 579.1M stations
# 1.2G aggregated
# 6.0G normalized
# 37.2G csvsScan a local directory:
disk-tree index /Users/ryan/c/disk-tree
# 97 files in 47 dirs, total size 1.5M
# 0B disk-tree/__init__.py
# 77B disk-tree/requirements.txt
# 867B disk-tree/setup.py
# 2.3K disk-tree/README.md
# 23.8K disk-tree/disk_tree
# 291.8K disk-tree/screenshots
# 580.4K disk-tree/.git
# 628.6K disk-tree/wwwdisk-tree caches scan results as Parquet files in ~/.config/disk-tree/scans/, with metadata in ~/.config/disk-tree/disk-tree.db. Override with DISK_TREE_ROOT.
- Local filesystems: Uses
gfind -printffor fast stat collection (handles sparse files correctly with 512-byte block sizes) - S3: Caches
aws s3 ls --recursiveoutput - Fresher child patching: When viewing a parent directory, newer child scans automatically patch in updated stats
- Depth-based predicate pushdown: Parquet queries filter by depth for fast loading
# Python
uv sync
disk-tree-server
# Web UI
cd ui
pnpm install
pnpm dev # http://localhost:5180 (proxies API to :5001)
# Screenshots
cd ui
pnpm screenshots