Feature matrix

Show/Hide columns (click on store/column name to toggle visibility):

Info columns: Category, Feature, Description, User Weight

Stores: VOS , Virtuoso Enterprise , 4store , Jena Fuseki HDT , Marklogic, GraphDB , Blazegraph , Allegrograph , fluidOps , Stardog , LDF ,

Category Feature Description User weight Virtuoso Open Source 7.2.4 Virtuoso Enterprise 7.2.4 4store Jena Fuseki (HDT) Marklogic Enterprise 8 GraphDB Cloud 7.0.1 Blazegraph 2.1 Allegrograph 6.1 fluidOps FedX 3.2 Stardog 4.2.1 LDF
RDF Store SPARQL 1.1 W3C spec SPARQL 1.1 documentation SPARQL 1.1 documentation Stackoverflow post: No yes (Jena Documentation) yes yes yes yes (documentation) Help:SPARQLEndpoint : yes yes BGP queries only
Database CRUD CRUD yes yes yes (SPARQL operators and update endpoint) HDT is read-only yes yes yes yes FluidOps supports local endpoint with CRUD, FedX is federator yes HDT is read-only
RDF Store Nested SELECT yes yes No yes yes (example query) yes yes (+ extension named subquery) yes yes (tested) yes? No
RDF Store Property Paths W3C Property paths Property paths in Virtuoso Property paths in Virtuoso No yes Property path expressions yes yes yes ? yes No
Database Transactions ACID, BASE ACID ACID No ACID ACID ACID ACID ACID FedX is not a database, individual endpoints can be chosen at will ACID Interface is read-only
Database Full-text Search Full-text-search yes, proprietary (bif:contains) yes, proprietary (bif:contains) no, only SPARQL regex Lucene/SOLR integration yes, built-in Lucene Both internal as SOLR SOLR integration Help:Search: yes (Lucene) Built-in based on Lucene Research effort
Database Storage layout Quad Table Quad Table ? Compressed triple file Multi-model: natively stores XML, JSON and RDF triples Not specified, changes between versions Not specified Table (5 columns) FedX is not a database, individual endpoints can be chosen at will Not specified Individual endpoints can be chosen at will, currently mainly HDT
Database Versioning yes, optional yes, optional No No yes, optional No? No? No? FedX is not a database, individual endpoints can be chosen at will yes,optional " under development (OSTRICH)
RDF Store Provenance Reification, Named Graphs Graph URI Graph URI Graph URI Graph URI , Reification Graph URI Reification and Graph URI Graph URI FedX is not a database, individual endpoints can be chosen at will Graph URI No
RDF Store Inference Support for reasoning RDFS & OWL RDFS & OWL No RDFS & OWL RDFS & OWL RDFS & OWL RFDS & OWL subset of RDFS & OWL Inferencing flag RDFS & OWL No
Usability (Open) Licensing Open Source, GPLv2 Commercial license for cluster edition Open Source GPLv3 Jena:Apache v2 , HDT: GNU v3 commercial with limited developer edition (1 B triples, single node) commercial with limited developer edition (max 2 concurrent clients) Open Source (+ extended commercial license) commercial, limited developer edition (50 M triples) FedX is Open Source: AGPL commercial, limited community edition (25 M triples) MIT license
Performance Clustering Techopedia - Clustering no yes yes No (Elephas project is beta) yes yes yes Allegrograph Federation Mechanism FedX = Federation yes Supports Federated Querying
Usability Development Ongoing yes yes Website went offline no yes yes Release notes Recently added functionality => yes yes Release 2017 recent releases
Usability Web UI (GUI) Conductor Web UI Conductor Web UI yes Limited GUI Admin UI GraphDB Workbench Blazegraph Workbench AGWebView fluidOps Information Workbench Stardog Web Console ldf-client browser
Usability Command Line interface (CLI) yes, isql yes, isql limited set of routines limited set of routines yes, query console No No yes, agquery FedX CLI yes, stardog & stardog admin limited set of routines
Usability Release Stable (Commercial) yes yes Offline No, research yes yes yes yes yes yes no, research
Usability Enterprise Support Not Included, support forums Web-based (free), Support contract No No Support Contract Email support Mailing list, issue tracker Paid Technical Support Free mailinglist, Support Contract Enterprise edition: phone, mail, community not No
Usability REST API yes yes No yes (Fuseki) yes yes yes yes Help:RestAPI => yes yes HTTP-client
Performance Bulk Loader Virtuoso Bulk RDF Loader Virtuoso Bulk RDF Loader File ingest HDT tools docker Marklogic content pump via workbench => local file ingest from server Bulk Data Loader Allegrograph Bulk Loader (agload) FedX is not a database, individual endpoints can be chosen at will yes HDT tools docker
Information Getting Started Getting Started Getting Started Website went offline -> backup HDT integration with Jena Getting Started with Marklogic Quick-start Quick-start Quick-start Help:GettingStarted Quick start guide LDF software overview
Information Source Code Github closed source Garlik/4store HDT-fuseki source closed source closed source Blazegraph Github closed source Not available closed source Github source code
Information Largest Reported Benchmark 15.4B 150B (8 nodes, 256GB RAM) Garlik internal usage: 15 BT HDT datasets - Freebase (2.06 BT), Note: multiple hdt graphs per store possible! ? UNIPROT : - close to 13 billion triples results for BigData: 12.7B > 1000B LargeRDFBench: > 1B 50B >1B (this work)
Information Technical Documentation Documentation Documentation Github hdt-fuseki Product Documentation GraphDB documentation Wiki Documentation IWB User's guide Quick Start Guide TPF client README
Information Cluster Installation Guide No Cluster Installation Website went offline - copy No Cluster configuration Single-Master setup guide Cluster Guide Replication Help:FedX HA Cluster Installation LDF
Usability Public Docker instance tenforce/virtuoso No inutano/4store Jena Fuseki but not HDT, HDT tools docker Marklogic Documentation No (commercial) Not latest version franzinc/agraph No bluepeppers (license key needed!) server.js and client.js
Information Download Link VOS Downloads No Website went offline - copy Maven build Developer version GraphDB Free Sourceforge link Download page IWB download page Request download link clients and server implementations
Database Database Model Graph DBMS , RDF Store , Document Store, Search Engine , XML DB Document Store, RDF Store, XML DB, Relational DB Document Store, RDF Store, XML DB, Relational DB RDF store RDF store Document Store, RDF Store, Search engine, XML DB GraphDB Graph DBMS, RDF Store Graph DBMS, RDF Store FedX is not a database, individual endpoints can be chosen at will Graph DBMS, RDF Store RDF Store
Information Initial Release 2006 1998 2009 HDT (2013) 2001 2000 2016 2004 FedX paper: 2011 2010 2014
Information Latest Version 2016 2016 2011 2016 2015 June 2016 2016 2015 FedX 3.1 2016 2016 2016
Information Implementation Language C C C hdt-cpp (C++), hdt-java, jena (java) C++ Java Java Common Lisp FedX.jar => Java Java Java, javascript
Usability Server Operating Systems AIX, FreeBSD, HP-UX, Linux, OS X, Solaris Windows AIX, FreeBSD, HP-UX, Linux, OS X, Solaris Windows Linux, OS X, ? Linux, OS X, Windows Linux, OS X, Solaris, Windows Linux, OS X, Windows Linux, OS X, Windows Linux, OS X, Windows All OS with JVM All OS with JVM All OS with JVM
Database Data Scheme yes (Column store) yes (Column store) schema-free (native triple store) RDF (schema-free) schema-free schema-free schema-free RDF FedX is not a database, individual endpoints can be chosen at will schema-free LDF is an RDF query interface => RDF (schema-free)
Database Typing yes yes ? yes yes yes yes yes yes yes no, xsd supported but every URI treated as a string
Database XML Support XML processing, XPath, SQuery, XSLT yes yes No only (RDF/XML) yes no no no Help:TreeDataSources: XML and JSON no only (RDF/XML)
Database Secondary Indexes Primary vs Secondary yes yes ? No (only hdt index) yes yes yes yes FedX is not a database, individual endpoints can be chosen at will yes No (only hdt index)
Database SQL yes yes no no SQL92 no no no Help:PolyglotWidgets no no
Usability Programming Languages .Net, C, C#, C++, Java, Javascript, Perl, PHP, Python, Ruby, Visual Basic .Net, C, C#, C++, Java, Javascript, Perl, PHP, Python, Ruby, Visual Basic PHP, Ruby, Python, Java Jena = Java framework C, C#, C++, Java, Javascript, Perl, PHP, Python, Ruby .Net, C#, Clojure, Java, JavaScript (Node.js), PHP, Python, Ruby, Scala .Net, C, C++, Java, JavaScript, PHP, Python, Ruby C#, Clojure, Java, Lisp, Perl, Python, Ruby, Scala Java .Net, Clojure, Groovy, Java, JavaScript, Python, Ruby Clients: JavaScript, Perl, Java, Python, Sever: +Netkernel and PHP
Usability Server-side scriptings Stored Procedures yes yes No? No yes (javascript, Xquery) yes (Java server plugin) yes yes /Help:QueryStoredProcedures: yes no No
Database Triggers Database Triggers yes yes No? No yes no no yes Help:SPARQLTriggerProvider: yes yes No
Performance Partitioning Methods Only single-node Sharding Sharding Only single-node Sharding no Sharding no FedX = manual partitioning no Manual partitioning (federation)
Performance Replication Methods Single-node Master-master replication and Master-slave Replicated segments No yes yes yes Master-slave replication No Master-master replication No replication
Database MapReduce no no No No (Elephas project is beta) yes no no no No no No
Database Consistency Type Immediate Immediate ? Immediate Immediate Immediate Consistency, Eventual consistency Immediate and Eventual possible Immediate and Eventual possible FedX is not a database, individual endpoints can be chosen at will Immediate No consistency (HDT)
Performance Concurrency Concurrency control yes yes No yes yes yes yes yes yes yes yes
Database Durability Durability yes yes yes yes (HDT = on disk) yes yes yes yes yes yes yes (HDT = on disk)
Database Access Control fine grained access rights according to SQL-standard fine grained access rights according to SQL-standard ? Access control via Jena Security Role-based access control at the document level Role-based access to repositories Security and Authentication via Web Application Container (Tomcat, Jetty) Users with fine-grained authorization concept, user roles and pluggable authentication Help:AccessControl => ACLs (role-based) Access rights for users and roles No
Usability Cloud Image Digital Cube Reseller Virtuoso PAGO ami No No Marklogic PAGO ami Ontotext PAGO ami No Allegrograph Community ami No Graviton to build Stardog image (requires license) No
RDF Store TPF Interface LDF in depth No No No No No No Collaboration with Olaf Hartig No No No Yes, native
Database Graph Traversal functionality Graph Traversal Limited: SQL extension for transitive traversal Limited: SQL extension for transitive traversal No No No No GAS API and Tinkerpop3 (Gremlin) Javascript-based interface for graph traversal No Tinkerpop3 (Gremlin) No
Database DBE Ranking Ranking 2 2 9 3 1 6 11 5 No 7 No

NOTE: Dereferencing URIs for FluidOps won't work, since they are for localhost:50080 and require installing FluidOps workbench

Scores per RDF Storage Solution

Store Points Database Points Performance Points RDF Store Points Usability Total Score
VOS 0 0 0 0 0
Virtuoso Enterprise 0 0 0 0 0
4store 0 0 0 0 0
Jena Fuseki HDT 0 0 0 0 0
Marklogic 0 0 0 0 0
GraphDB 0 0 0 0 0
Blazegraph 0 0 0 0 0
Allegrograph 0 0 0 0 0
fluidOps 0 0 0 0 0
Stardog 0 0 0 0 0
LDF 0 0 0 0 0