July 2026 Summaries

In a detailed comparison of graph traversal performance between PostgreSQL 19 and Memgraph, the focus is on evaluating the effectiveness of SQL/PGQ support in PostgreSQL for exact hop reachability queries. PostgreSQL 19 introduces SQL/PGQ, allowing property graph queries over relational tables, which is valuable for teams already using Postgres to explore connected data. However, as traversal depth increases, PostgreSQL's performance diminishes, with significant slowdowns evident at four to five hops, where Memgraph, a native graph database built around the property graph model, demonstrates superior efficiency. Memgraph's architecture, which treats nodes and relationships as core data models, enables it to maintain low latency even in deep traversals, making it more suitable for applications that rely heavily on graph-based operations. The benchmark, utilizing the Pokec dataset, highlights the strengths and limitations of each database, suggesting that while PostgreSQL 19's graph query capabilities are beneficial for shallow queries within existing relational data, Memgraph is more advantageous for workloads requiring extensive graph traversal.

Memgraph 3.11 enhances multi-tenancy for cross-database graph workloads by simplifying the management and operation of graph data across multiple databases. This release introduces a cross_database module that facilitates querying across databases without external stitching, reducing the complexity of cross-boundary workflows and enabling seamless data migration and environment comparison. To address the need for temporary data views, Memgraph 3.11 offers virtual graph views through derive(), allowing users to explore graph structures without creating permanent data storage. Additionally, the update improves visibility and security in multi-tenant systems with per-database labels for monitoring tools like Prometheus and Grafana, enhanced query logging, and intra-cluster TLS for secure communication. These features collectively make it easier to manage, monitor, and secure multi-tenant graph workloads, providing teams with practical tools to handle the increasing complexity of connected graph data systems while maintaining necessary data isolation and control.