System
GNOME Tracker Cache (Legacy)
On older GNOME desktops, Tracker stores search and metadata indexing cache under ~/.cache/tracker to speed up file discovery, thumbnail lookup, and SPARQL query results. That cache can include extracted metadata blobs, ontology and query caches, thumbnail references, and temporary database files created while indexing files, emails, and media. Kudu removes only this legacy Tracker cache so stale index artifacts and temporary data can be rebuilt, without touching your documents, account data, or desktop settings.
Why clean GNOME Tracker Cache (Legacy)?
- Stale metadata extracted from renamed or moved files can leave GNOME search showing wrong titles, tags, or duplicate results even after the files themselves changed
- Corrupted Tracker cache files can make Nautilus search return incomplete results or miss recently added documents until the indexer rebuilds
- Leftover temporary database files from interrupted indexing runs can keep tracker-miner busy in the background, which shows up as constant disk activity and higher CPU use when the system is idle
- Outdated thumbnail and media metadata references can cause old previews or incorrect duration and codec details to appear for videos and music in GNOME apps
- A cache built against older ontology data or miner state can trigger repeated reindexing, so users notice slow logins, noisy disk access, or search results that keep changing
- SQLite page fragmentation in Tracker's legacy databases wastes space and slows metadata lookups; VACUUM rewrites the database file to compact pages without deleting rows
- Large accumulated cache data under ~/.cache/tracker can consume hundreds of megabytes over time, especially on systems with many indexed files, and cleaning it frees space without removing the actual files
What gets cleaned
Cache paths Kudu targets
Linux
~/.cache/tracker |
Frequently asked
Common questions about GNOME Tracker Cache (Legacy)
Related cleaners
Free & open source
Download Kudu and reclaim your disk space.
Available on Windows, macOS, and Linux. No account required, no feature gates, no telemetry without consent. All cleaning targets are open source and community-auditable.