github.devotd.
NewData

apache/age PG18/v1.8.0-rc0 — openCypher extension for PostgreSQL

A rundown of apache/age, which provides a graph data model and openCypher queries as a PostgreSQL extension, and of the status of the PG18/v1.8.0-rc0 release candidate.

Repositoryapache/age ↗

A rundown of apache/age, which provides a graph data model and openCypher queries as a PostgreSQL extension, and of the status of the PG18/v1.8.0-rc0 release candidate.

GitHub description: Graph database optimized for fast analysis and real-time data processing. It is provided as an extension to PostgreSQL.

stars
★ 4.9K
Forks
538
Language
C
License
Apache-2.0
Latest release
PG18/v1.8.0-rc0
Read original source ↗View on GitHub ↗

apache/age is a graph database aimed at fast analytics and real-time data processing, distributed as a PostgreSQL extension. Rather than running a separate server, it stores graph data inside an existing PostgreSQL instance and queries it with the graph query language openCypher. The most recent release on record is the candidate-stage PG18/v1.8.0-rc0. Because the given record contains no changelog, what changed in this release cannot be known.

The graph data model draws relationships as vertices and edges. The README states that a single repository handles both relational and graph data, supporting ANSI SQL and openCypher alike. It also offers queries that mix SQL and Cypher, queries that handle multiple graphs, hierarchical graph label configurations, and property indexes on vertices and edges. The implementation language is C, the license is Apache-2.0, the star count is 4907, and the public release was in July 2020.

That said, the momentum score in the research record is a low 2.9 out of 10, and the novelty, technology, and quality scores are not stated. The given information contains no performance figures backing the claim of targeting fast analytics and real-time data processing, and no comparison results against other graph databases. With so little hard evidence, it is more a project worth keeping as a candidate for when you already use PostgreSQL and need a graph model for only some queries than one to adopt right now.