Release: AllegroGraph 4.0, 100% ACID
AllegroGraph RDFStore[1], a solution at the crossing of RDF stores[2] and graph databases, has released recently a new major update featuring:
- AllegroGraph is 100 percent ACID, supporting Transactions: Commit, Rollback, and Checkpointing. See the new tutorials for the Java and Python clients
- Full and Fast Recoverability
- 100% Read Concurrency, Near Full Write Concurrency
- Online Backups
- Dynamic and Automatic Indexing – All committed triples are always indexed (7 indices)
- Advanced Text Indexing – Lucene style but faster, text indexing per predicate. See the new tutorials for the Java and Python clients
- Duplicate Triple deletion while indexing
- All Clients based on http REST Protocol – Java, Sesame, Jena, and Python
- Completely multi-processing based (SMP) – Automatic Resource Management for all processors and disks, and optimized memory use. See the new performance tuning guide here, and new server configuration guide here
- Column-based compression of indices similar to column-based RDBMS – reduced paging, better performance
- Dedicated and Public Sessions – In dedicated sessions users can work with their own rule sets against the same database
- Python Client Improvements – We now provide a full Python interface. The API is based on the Java Sesame interface and includes Spatial-Temporal and Social Network support
- LUBM Benchmarks – Updated for this release
As a side note, in the graph databases space, AllegroGraph is not the first one supporting ACID, Neo4j being fully transactional for quite a while[3].
