Docker compose单实例部署Milvus检索速度过慢

[nlp-greyfoss]

参考官方文档，Docker compose单实例部署Milvus检索速度过慢。

version: '3.5'

services:
  etcd:
    container_name: milvus-etcd
    image: quay.io/coreos/etcd:v3.5.5
    environment:
      - ETCD_AUTO_COMPACTION_MODE=revision
      - ETCD_AUTO_COMPACTION_RETENTION=1000
      - ETCD_QUOTA_BACKEND_BYTES=4294967296
      - ETCD_SNAPSHOT_COUNT=50000
    volumes:
      - ${DOCKER_VOLUME_DIRECTORY:-.}/volumes/etcd:/etcd
    command: etcd -advertise-client-urls=http://127.0.0.1:2379 -listen-client-urls http://0.0.0.0:2379 --data-dir /etcd
    healthcheck:
      test: ["CMD", "etcdctl", "endpoint", "health"]
      interval: 30s
      timeout: 20s
      retries: 3

  minio:
    container_name: milvus-minio
    image: minio/minio:RELEASE.2023-03-20T20-16-18Z
    environment:
      MINIO_ACCESS_KEY: minioadmin
      MINIO_SECRET_KEY: minioadmin
    ports:
      - "9001:9001"
      - "9000:9000"
    volumes:
      - ${DOCKER_VOLUME_DIRECTORY:-.}/volumes/minio:/minio_data
    command: minio server /minio_data --console-address ":9001"
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:9000/minio/health/live"]
      interval: 30s
      timeout: 20s
      retries: 3

  standalone:
    container_name: milvus-standalone
    image: milvusdb/milvus:v2.4.13-hotfix
    command: ["milvus", "run", "standalone"]
    security_opt:
    - seccomp:unconfined
    environment:
      ETCD_ENDPOINTS: etcd:2379
      MINIO_ADDRESS: minio:9000
    volumes:
      - ${DOCKER_VOLUME_DIRECTORY:-.}/volumes/milvus:/var/lib/milvus
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:9091/healthz"]
      interval: 30s
      start_period: 90s
      timeout: 20s
      retries: 3
    ports:
      - "19530:19530"
      - "9091:9091"
    depends_on:
      - "etcd"
      - "minio"

networks:
  default:
    name: milvus

milvus:v2.4.13-hotfix

按500token分块，共260+个文本块，数量不多：

但是检索耗时比期望高，代码如下：

from pymilvus import (
    AnnSearchRequest,
    WeightedRanker,
)


def dense_search(col, query_dense_embedding, limit=10):
    search_params = {"metric_type": "IP", "params": {}}
    res = collection.search(
        [query_dense_embedding],
        anns_field="dense_vector",
        limit=limit,
        output_fields=["text"],
        param=search_params,
    )[0]
    return [hit.get("text") for hit in res]

%%time
# embed_query 耗时 100ms
dense_query_emb = dense_embedding.embed_query("收入证明怎么开")
dense_search(collection, dense_query_emb)

1it [00:00, 24.38it/s]
CPU times: user 9.53 ms, sys: 4.86 ms, total: 14.4 ms
Wall time: 526 ms

[yhmo]

估计那个collection的consistency level是Strong.
可以试下search里加一行参数设置consistency_level：

res = collection.search(
        [query_dense_embedding],
        anns_field="dense_vector",
        limit=limit,
        output_fields=["text"],
        param=search_params,
        consistency_level="Bounded"
    )[0]

[yhmo]

在langchian.Milvus对象的similarity_search接口里加。
similarity_search接口由基类VectorStore定义：

def similarity_search(
        self, query: str, k: int = 4, **kwargs: Any
    )

那个"**kwargs"能把额外的参数传给里面的collection.search()

[nlp-greyfoss]

谢谢，我知道怎么解决了。创建Collection的时候指定就好了，后续检索的时候不用传。

[nlp-greyfoss]

非常感谢，确实变快了，但是单机版为什么还会受这个字段影响。

[xiaofan-luan]

这个是因为milvus的同步机制的导致的。数据写入日志后不会等apply。

对于强一致查询，需要等至少一个同步信号，同步信号的同步周期是200ms，也就是说任何查询至少要200-400ms之后才能执行