Inspecting Files
Disk Usage
Crab stores data in several places: hydrated files in your working tree, chunks in the staging area, downloaded data in the local cache, and permanent storage in the cloud. Understanding this breakdown helps you decide when to dehydrate, prune, or clean.
Storage Layers
Checking Disk Usage
crab duWorking tree:
Tracked files: 42
Hydrated: 12 files (3.2 GB)
Pointers: 30 files (4.5 KB)
Staging: 45.2 MB
Local cache: 1.8 GB
Total local: 5.0 GBInclude remote storage
crab du --remoteAdds the cloud storage size (requires network access):
Remote: 8.4 GB (crab://my-bucket/my-repo)Understanding the Numbers
| Layer | What's stored | How to reduce |
|---|---|---|
| Working tree (hydrated) | Full file content on disk | crab dehydrate |
| Working tree (pointers) | Tiny pointer stubs (~128 B each) | Nothing needed |
| Staging | Chunks waiting to be pushed | crab staging clean after push |
| Local cache | Previously downloaded chunks | crab prune or crab cache clean |
| Remote | All pushed xorbs and metadata | crab gc for unreachable objects |
Reclaiming Space
Quick wins
# Free the most space: dehydrate files you're not using
crab dehydrate --all
# Clean staging after a successful push
crab staging clean
# Remove unreferenced cache entries
crab pruneCheck what you'd save
# See how much hydrated content you have
crab du
# See what GC would reclaim remotely
crab gc --dry-runCLI Reference
For complete command syntax and JSON output format, see the crab du reference.