One of the most significant fixes in this version involves memory pressure during large-scale data ingestion. Users previously reported occasional OOM (Out of Memory) errors when importing massive CSV or Parquet files into a graph schema.
. As an embeddable, serverless graph database management system designed for speed, scalability, and complex analytical workloads, Kùzu has become a staple for building AI applications like GraphRAG and managing local knowledge graphs.
The core lesson in both cases is the same: thorough, step-by-step troubleshooting works best. Whether you're checking your Kùzu installation version or your Chevy Cruze battery cables, starting with the simplest potential fix often saves the most time. kuzu v0 136 fixed
Eliminated redundant and unused getEstimatedMemUsage function calls across the query graph.
In-memory and embedded databases must manage RAM with extreme precision. Previous iterations occasionally suffered from edge-case memory leaks during highly complex recursive joins or massive MERGE operations. One of the most significant fixes in this
Kùzu utilizes vectorized and factorized query processing to parallelize execution across multi-core systems.
that occurred during intensive and concurrent query processing. Cypher Parser Robustness
CSV bulk load example (place CSVs in accessible dir):
However, version 0.135 introduced several regressions that hampered production use. The core issues ranged from race conditions in multi-threaded environments to a persistent segmentation fault when parsing certain data structures. The community has been eagerly awaiting a stable release, and with , those prayers have been answered.
The "fixed" aspect of version 0.1.3.6 focuses on three main pillars: , Cypher Parser Robustness , and Storage Layer Consistency . 1. Improved Memory Handling during Bulk Loads
The fixes in this version contribute to a more production-ready embedded graph database.