Example layers written in Python for FoundationDB. These layers require the FoundationDB download to work.

FoundationDB decouples its data storage technology from its data model. Layers add capabilities to FoundationDB's ordered key-value API. A layer can provide a new data model, compatibility with existing systems, or even serve as an entire framework.

These layers demonstrate the concept of layer development: that even simple layers inherit the transactional, scalable, fault-tolerant properties of FoundationDB. While functional, these layers are designed primarily to be simple, clear examples.

• blob.py - Arbitrary-sized and sparse large binary objects.
• bulk.py - Bulk-loads external datasets to FoundationDB with extensible support for CSV, JSV, and blobs.
• counter.py - High-performance counter that illustrates the use of dynamic sharding for high contention conditions. Note: This layer was implemented prior to the release of our atomic operations. Counters can now be more effectively implemented using an atomic operation.
• directory.py - Directories for administering layers and their respective subspaces. Directories are identified by paths analogous to the paths in a Unix-like file system.
• pubsub.py - Message passing according to the publish-subscribe pattern. Allows management of feeds and inboxes as well as message delivery.
• queue.py - Queues supporting a high contention mode for multiple clients and an optimized mode for single clients.
• simpledoc.py - A simple, hierarchical data model for storing document-oriented data. Supports a powerful plugin capability with indexes.
• spatial.py - A spatial index for 2D points that allows efficient queries of axis-aligned rectangular regions. Does dimensionality reduction via a Z-order fractal curve (aka geohash).
• stringintern.py - For interning (aka normalizing, aliasing) commonly-used long strings into shorter representations. Maintains the normalization state in the database, as well as a local cache for high performance.
• subspace.py - Subspaces of keys. Should be used to manage namespaces for application data.
• vector.py - Vectors for storing and manipulating potentially sparse arrays.

Discussion of layers and layer development takes place on the FoundationDB community site. (Sign up here.)