Datanet is an open source CRDT based data synchronization system
- enables ubiquitous write-thru caching
- gracefully tolerates multiple levels of system failures
- replicates data in a P2P/asynchronous/zero-coordination flow
- scales by design
- supports advanced data-structures
- has a simple API: JSON



Datanet aims to achieve ubiquitous write though caching. CRDT replication can be added to any cache in your stack, meaning data modifications to these stacks are globally & reliably replicated. Locally modifying data yields massive gains in latency, produces a more efficient replication stream, & is extremely robust. It’s time to prefetch data to compute.




Optimal Replication Latency:
Agents replicate lazily, outside the fast path, in a P2P mesh. Modifications are applied immediately (single phase distributed commit) without sacrificing strong data guarantees.
Scalable/Elastic/Fault-tolerant on a new level:
Write availability for temporarily offline Agents, Elastic highly-available geo-distributed mesh of central clusters, Survive multiple simultaneous DataCenter failures ... seamlessly.


CRDT GarbageCollection:
Distributed, asynchronous, zero-coordination. Datanet's GC algorithm is itself a CRDT algorithm: same strong guarantees & optimized latency. GC opens up new types of CRDTs.
Simple API + Advanced Data Structures:
Supports ALL JSON data-structures [numbers, strings, dictionaries, sequences] plus extensible data structures [Lists, Trees, Queues, Stacks, etc...].