← All results

Vector

native · observability-pipeline · MPL-2.0 · active · rust

Homepage · Source · Its configuration in this repository

Has this configuration been reviewed upstream?

Not yet. This configuration has not been checked by the people who maintain this system, and until it has, treat its numbers as ours rather than theirs.

What it was asked to guarantee

at-least-once · offset-commit every 5s

The envelope it ran in

32 CPU · 96g memory

• sut — data-plane · 32 CPU · 96g

Its arms

• JSONEachRow json-each-row · realistic · json_each_row · default
  threads 32 · chunk_size_events 1000 · sources 8 · batch_events 262144 · batch_timeout_secs 1 · request_concurrency 32 · buffer_events 524288 · compression none

Where it has been measured

Configurations it refuses

• buffer_events must exceed batch_events. The sink's memory buffer feeds the batcher, so it must hold strictly more than one batch: smaller, and every batch seals early on buffer starvation; equal, and the next batch cannot begin filling while the last drains — either way the declared batch_events is not the quantity actually in force and the record would misdescribe the arm. Raise buffer_events with batch_events.

Declared deviations

• Static inline Avro schema with `strip_schema_id_prefix`; the Schema Registry is never contacted. The 5-byte Confluent frame is stripped without validating the schema id, no registry lookup is ever paid, and the arm cannot detect a writer-schema change mid-run. — Vector has no Schema Registry integration. The baked schema is manufactured at image build from the committed workload/schema/sensor_batch.avsc, so it cannot drift from the schema every other arm fetches. · affects decode-path

Tell us we got this wrong

If this system is configured badly here, that is a bug in this benchmark rather than a result about Vector, and the pull request that fixes it is the most valuable one this repository can receive. How to send one. The whole configuration is at entrants/vector.

This system also reports CPU per row, Cores, data plane, Peak memory, data plane, Peak charged memory, Throttled, ClickHouse CPU per row, ClickHouse CPU, ClickHouse CPU per written row, ClickHouse CPU wait, Bytes inserted, Rows per insert, Rows written server-side. Those figures are on each arm’s own disclosure in the results table.