2026.09.14
Urocissa: A Self-Hosted Gallery for Massive Photo Libraries
Urocissa is a Rust and Vue self-hosted gallery for large photo and video collections. It offers search, tags, EXIF, virtual scrolling, and Docker deployment, while multi-user accounts, face recognition, and existing-folder support are listed as unavailable.
Project focus
Urocissa is aimed at people who want to manage large photo and video collections on their own infrastructure. Built with Rust and Vue, it focuses on browsing, searching, and filtering very large galleries without loading the entire database into the browser. The README states a goal of serving one million photos on a 4 GB RAM server, but that should be read as a project goal and estimate rather than a guarantee for every deployment.
Documented capabilities
The README lists common gallery-management features: photo and video uploads, downloads, EXIF data, user-defined tags, duplicate handling, timeline navigation, instant select-all, responsive layout, shareable albums, and Docker installation. Search supports boolean operators such as and, or, and not. Its infinite photo stream uses virtual scrolling to reduce the effect of very large collections on browser layout limits.
Urocissa uses an in-memory cached database for access and search. The README includes RAM estimates by photo count and explicitly says that the benchmark covers metadata-only records, not the media files themselves. Larger figures are linear extrapolations, and actual use may vary with allocator behavior, indexing options, and memory reuse. Operators therefore need additional capacity for original files, thumbnails, the operating system, containers, and other services; the table should not be treated as a complete machine-sizing guide.
Who may benefit
The project may suit users with extensive personal media who want self-hosted organization and fast retrieval. Linux users familiar with Docker have a documented container-based setup path. The README also provides download or build guidance for Windows and Linux. People with small collections, teams requiring mature account-based collaboration, or users dependent on mobile applications should compare those needs with the documented feature set before adopting it.
Adoption limits
The supplied project material explicitly lists multi-user support, object or face recognition, geolocation and maps, an Android app, external libraries, and existing-folder support as unavailable. Discovery is marked as in progress. This means adopters should not assume that Urocissa can directly index an existing directory, recognize faces, or provide a mobile app and team-account model.
The website uses privacy and security-oriented language, but the supplied sources do not provide enough evidence for an independent security conclusion. Anyone exposing the service publicly should separately plan authentication, network boundaries, backups, and access control. Release 3.4.1 addresses a large-gallery virtual-scrolling issue by keeping the physical browser buffer bounded instead of applying the full logical collection height to rendered layout. Checking formal releases and documentation is preferable to judging stability from a demo alone. The README says the demos are read-only, so they do not represent upload, editing, or permission workflows.
Deployment considerations
Docker is the clearly documented quick-start option, with configuration documentation covering items such as ports and passwords. Before production use, confirm media storage, backup and restore procedures, upgrade rollback, and external exposure. Those operational concerns are not replaced by the memory estimates. If the library approaches a million records, test with your own metadata, indexing choices, and concurrent-access pattern to verify remaining memory headroom.
Sources
Sources checked: 2026-09-14