Skip to content

Better benchmarking with JMH for time series data ingest #51

Description

@wmehling

*Task: Convert and extend the existing KVIN service benchmark into a structured Java Microbenchmark Harness (JMH) suite.

The suite should cover:

  • the complete server-side ingestion path
  • request and payload parsing
  • SPARQL/RDF-to-KVIN conversion
  • KVIN encoding, ID resolution, and batching
  • KvinLevelDb and raw LevelDB write baselines
  • different batch sizes, URI cardinalities, and concurrency levels

Record throughput, latency, allocation/GC metrics, and profiler data where useful.

The goal is to identify the limiting stage in the server-side ingestion pipeline and quantify the overhead of each LinkedFactory layer.

With the knowledge of what limits throughput, we can develop ideas on how to improve Writing Speed.

Limits: This is about streaming / writing to the LF only.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

enhancementNew feature or request

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions