search

qdrant / vector-db-benchmark

Framework for benchmarking vector search engines
https://qdrant.tech/benchmarks/
Apache License 2.0
270 stars 77 forks source link

Cherry picking some milvus improvements. Updated milvus client to include backoff strategy on ingestion. Added extra configurations to ease remote benchmark run. #170

Open filipecosta90 opened 2 months ago

filipecosta90 commented 2 months ago

Fixes #54 Furthermore, it also eases the process of running the milvus benchmark against remote hosts (like zilliz cloud deployments, or other SW deployments ).

This has been used on our benchmarks, and was part of a bigger work in https://github.com/redis-performance/vector-db-benchmark.

Example of running targeting a zilliz cloud cluster:

MILVUS_USER="db_admin" \
MILVUS_PASS="<...>" \
MILVUS_PORT=<zilliz port> \
          python3 run.py  --engines milvus-m-* --datasets gist-960-euclidean \
                          --host https://<endpoint info>.zillizcloud.com:<zilliz port>