vectordb-bench 0.0.12__tar.gz → 0.0.13__tar.gz

{vectordb_bench-0.0.12 → vectordb_bench-0.0.13}/.gitignore

@@ -9,3 +9,4 @@ __MACOSX
 build/
 venv/
 .idea/
+results/

{vectordb_bench-0.0.12 → vectordb_bench-0.0.13}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vectordb-bench
-Version: 0.0.12
+Version: 0.0.13
 Summary: VectorDBBench is not just an offering of benchmark results for mainstream vector databases and cloud services, it's your go-to tool for the ultimate performance and cost-effectiveness comparison. Designed with ease-of-use in mind, VectorDBBench is devised to help users, even non-professionals, reproduce results or test new systems, making the hunt for the optimal choice amongst a plethora of cloud services and open-source vector databases a breeze.
 Author-email: XuanYang-cn <xuan.yang@zilliz.com>
 Project-URL: repository, https://github.com/zilliztech/VectorDBBench
@@ -36,10 +36,10 @@ Requires-Dist: pinecone-client; extra == "all"
 Requires-Dist: weaviate-client; extra == "all"
 Requires-Dist: elasticsearch; extra == "all"
 Requires-Dist: pgvector; extra == "all"
+Requires-Dist: pgvecto_rs[psycopg3]>=0.2.1; extra == "all"
 Requires-Dist: sqlalchemy; extra == "all"
 Requires-Dist: redis; extra == "all"
 Requires-Dist: chromadb; extra == "all"
-Requires-Dist: psycopg2; extra == "all"
 Requires-Dist: psycopg; extra == "all"
 Requires-Dist: psycopg-binary; extra == "all"
 Requires-Dist: opensearch-dsl==2.1.0; extra == "all"
@@ -56,10 +56,16 @@ Provides-Extra: pgvector
 Requires-Dist: psycopg; extra == "pgvector"
 Requires-Dist: psycopg-binary; extra == "pgvector"
 Requires-Dist: pgvector; extra == "pgvector"
+Provides-Extra: pgvectorscale
+Requires-Dist: psycopg; extra == "pgvectorscale"
+Requires-Dist: psycopg-binary; extra == "pgvectorscale"
+Requires-Dist: pgvector; extra == "pgvectorscale"
 Provides-Extra: pgvecto-rs
-Requires-Dist: psycopg2; extra == "pgvecto-rs"
+Requires-Dist: pgvecto_rs[psycopg3]>=0.2.1; extra == "pgvecto-rs"
 Provides-Extra: redis
 Requires-Dist: redis; extra == "redis"
+Provides-Extra: memorydb
+Requires-Dist: memorydb; extra == "memorydb"
 Provides-Extra: chromadb
 Requires-Dist: chromadb; extra == "chromadb"
 Provides-Extra: awsopensearch
@@ -105,7 +111,9 @@ All the database client supported
 | elastic                  | `pip install vectordb-bench[elastic]`       |
 | pgvector                 | `pip install vectordb-bench[pgvector]`      |
 | pgvecto.rs               | `pip install vectordb-bench[pgvecto_rs]`    |
+| pgvectorscale            | `pip install vectordb-bench[pgvectorscale]` |
 | redis                    | `pip install vectordb-bench[redis]`         |
+| memorydb                 | `pip install vectordb-bench[memorydb]`      |
 | chromadb                 | `pip install vectordb-bench[chromadb]`      |
 | awsopensearch            | `pip install vectordb-bench[awsopensearch]` |
 

{vectordb_bench-0.0.12 → vectordb_bench-0.0.13}/README.md

@@ -37,7 +37,9 @@ All the database client supported
 | elastic                  | `pip install vectordb-bench[elastic]`       |
 | pgvector                 | `pip install vectordb-bench[pgvector]`      |
 | pgvecto.rs               | `pip install vectordb-bench[pgvecto_rs]`    |
+| pgvectorscale            | `pip install vectordb-bench[pgvectorscale]` |
 | redis                    | `pip install vectordb-bench[redis]`         |
+| memorydb                 | `pip install vectordb-bench[memorydb]`      |
 | chromadb                 | `pip install vectordb-bench[chromadb]`      |
 | awsopensearch            | `pip install vectordb-bench[awsopensearch]` |
 

{vectordb_bench-0.0.12 → vectordb_bench-0.0.13}/install/requirements_py3.11.txt

@@ -5,6 +5,7 @@ pinecone-client
 weaviate-client
 elasticsearch
 pgvector
+pgvecto_rs[psycopg3]>=0.2.1
 sqlalchemy
 redis
 chromadb

{vectordb_bench-0.0.12 → vectordb_bench-0.0.13}/pyproject.toml

@@ -56,10 +56,10 @@ all = [
     "weaviate-client",
     "elasticsearch",
     "pgvector",
+    "pgvecto_rs[psycopg3]>=0.2.1",
     "sqlalchemy",
     "redis",
     "chromadb",
-    "psycopg2",
     "psycopg",
     "psycopg-binary",
     "opensearch-dsl==2.1.0",
@@ -71,8 +71,10 @@ pinecone = [ "pinecone-client" ]
 weaviate = [ "weaviate-client" ]
 elastic = [ "elasticsearch" ]
 pgvector = [ "psycopg", "psycopg-binary", "pgvector" ]
-pgvecto_rs = [ "psycopg2" ]
+pgvectorscale = [ "psycopg", "psycopg-binary", "pgvector" ]
+pgvecto_rs = [ "pgvecto_rs[psycopg3]>=0.2.1" ]
 redis = [ "redis" ]
+memorydb = [ "memorydb" ]
 chromadb = [ "chromadb" ]
 awsopensearch = [ "awsopensearch" ]
 zilliz_cloud = []

{vectordb_bench-0.0.12 → vectordb_bench-0.0.13}/vectordb_bench/backend/clients/__init__.py

@@ -30,7 +30,9 @@ class DB(Enum):
     WeaviateCloud = "WeaviateCloud"
     PgVector = "PgVector"
     PgVectoRS = "PgVectoRS"
+    PgVectorScale = "PgVectorScale"
     Redis = "Redis"
+    MemoryDB = "MemoryDB"
     Chroma = "Chroma"
     AWSOpenSearch = "OpenSearch"
     Test = "test"
@@ -70,10 +72,18 @@ class DB(Enum):
         if self == DB.PgVectoRS:
             from .pgvecto_rs.pgvecto_rs import PgVectoRS
             return PgVectoRS
+        
+        if self == DB.PgVectorScale:
+            from .pgvectorscale.pgvectorscale import PgVectorScale
+            return PgVectorScale
 
         if self == DB.Redis:
             from .redis.redis import Redis
             return Redis
+        
+        if self == DB.MemoryDB:
+            from .memorydb.memorydb import MemoryDB
+            return MemoryDB
 
         if self == DB.Chroma:
             from .chroma.chroma import ChromaClient
@@ -118,9 +128,17 @@ class DB(Enum):
             from .pgvecto_rs.config import PgVectoRSConfig
             return PgVectoRSConfig
 
+        if self == DB.PgVectorScale:
+            from .pgvectorscale.config import PgVectorScaleConfig
+            return PgVectorScaleConfig
+
         if self == DB.Redis:
             from .redis.config import RedisConfig
             return RedisConfig
+        
+        if self == DB.MemoryDB:
+            from .memorydb.config import MemoryDBConfig
+            return MemoryDBConfig
 
         if self == DB.Chroma:
             from .chroma.config import ChromaConfig
@@ -163,6 +181,10 @@ class DB(Enum):
             from .aws_opensearch.config import AWSOpenSearchIndexConfig
             return AWSOpenSearchIndexConfig
 
+        if self == DB.PgVectorScale:
+            from .pgvectorscale.config import _pgvectorscale_case_config
+            return _pgvectorscale_case_config.get(index_type)
+
         # DB.Pinecone, DB.Chroma, DB.Redis
         return EmptyDBCaseConfig
 

{vectordb_bench-0.0.12 → vectordb_bench-0.0.13}/vectordb_bench/backend/clients/api.py

@@ -15,6 +15,7 @@ class MetricType(str, Enum):
 class IndexType(str, Enum):
     HNSW = "HNSW"
     DISKANN = "DISKANN"
+    STREAMING_DISKANN = "DISKANN"
     IVFFlat = "IVF_FLAT"
     IVFSQ8 = "IVF_SQ8"
     Flat = "FLAT"
@@ -38,6 +39,22 @@ class DBConfig(ABC, BaseModel):
     """
 
     db_label: str = ""
+    version: str = ""
+    note: str = ""
+
+    @staticmethod
+    def common_short_configs() -> list[str]:
+        """
+        short input, such as `db_label`, `version`
+        """
+        return ["version", "db_label"]
+
+    @staticmethod
+    def common_long_configs() -> list[str]:
+        """
+        long input, such as `note`
+        """
+        return ["note"]
 
     @abstractmethod
     def to_dict(self) -> dict:
@@ -45,7 +62,10 @@ class DBConfig(ABC, BaseModel):
 
     @validator("*")
     def not_empty_field(cls, v, field):
-        if field.name == "db_label":
+        if (
+            field.name in cls.common_short_configs()
+            or field.name in cls.common_long_configs()
+        ):
             return v
         if not v and isinstance(v, (str, SecretStr)):
             raise ValueError("Empty string!")

vectordb_bench-0.0.13/vectordb_bench/backend/clients/memorydb/cli.py

@@ -0,0 +1,88 @@
+from typing import Annotated, TypedDict, Unpack
+
+import click
+from pydantic import SecretStr
+
+from ....cli.cli import (
+    CommonTypedDict,
+    HNSWFlavor2,
+    cli,
+    click_parameter_decorators_from_typed_dict,
+    run,
+)
+from .. import DB
+
+
+class MemoryDBTypedDict(TypedDict):
+    host: Annotated[
+        str, click.option("--host", type=str, help="Db host", required=True)
+    ]
+    password: Annotated[str, click.option("--password", type=str, help="Db password")]
+    port: Annotated[int, click.option("--port", type=int, default=6379, help="Db Port")]
+    ssl: Annotated[
+        bool,
+        click.option(
+            "--ssl/--no-ssl",
+            is_flag=True,
+            show_default=True,
+            default=True,
+            help="Enable or disable SSL for MemoryDB",
+        ),
+    ]
+    ssl_ca_certs: Annotated[
+        str,
+        click.option(
+            "--ssl-ca-certs",
+            show_default=True,
+            help="Path to certificate authority file to use for SSL",
+        ),
+    ]
+    cmd: Annotated[
+        bool,
+        click.option(
+            "--cmd",
+            is_flag=True,
+            show_default=True,
+            default=False,
+            help="Cluster Mode Disabled (CMD), use this flag when testing locally on a single node instance. In production, MemoryDB only supports cluster mode (CME)",
+        ),
+    ]
+    insert_batch_size: Annotated[
+        int,
+        click.option(
+            "--insert-batch-size",
+            type=int,
+            default=10,
+            help="Batch size for inserting data. Adjust this as needed, but don't make it too big",
+        ),
+    ]
+
+
+class MemoryDBHNSWTypedDict(CommonTypedDict, MemoryDBTypedDict, HNSWFlavor2):
+    ...
+
+
+@cli.command()
+@click_parameter_decorators_from_typed_dict(MemoryDBHNSWTypedDict)
+def MemoryDB(**parameters: Unpack[MemoryDBHNSWTypedDict]):
+    from .config import MemoryDBConfig, MemoryDBHNSWConfig
+
+    run(
+        db=DB.MemoryDB,
+        db_config=MemoryDBConfig(
+            db_label=parameters["db_label"],
+            password=SecretStr(parameters["password"]) if parameters["password"] else None,
+            host=SecretStr(parameters["host"]),
+            port=parameters["port"],
+            ssl=parameters["ssl"],
+            ssl_ca_certs=parameters["ssl_ca_certs"],
+            cmd=parameters["cmd"],
+        ),
+        db_case_config=MemoryDBHNSWConfig(
+            M=parameters["m"],
+            ef_construction=parameters["ef_construction"],
+            ef_runtime=parameters["ef_runtime"],
+            insert_batch_size=parameters["insert_batch_size"]
+        ),
+        **parameters,
+    )

vectordb_bench-0.0.13/vectordb_bench/backend/clients/memorydb/config.py

@@ -0,0 +1,54 @@
+from pydantic import BaseModel, SecretStr
+
+from ..api import DBCaseConfig, DBConfig, IndexType, MetricType
+
+
+class MemoryDBConfig(DBConfig):
+    host: SecretStr
+    password: SecretStr | None = None
+    port: int | None = None
+    ssl: bool | None = None
+    cmd: bool | None = None
+    ssl_ca_certs: str | None = None
+
+    def to_dict(self) -> dict:
+        return {
+            "host": self.host.get_secret_value(),
+            "port": self.port,
+            "password": self.password.get_secret_value() if self.password else None,
+            "ssl": self.ssl,
+            "cmd": self.cmd,
+            "ssl_ca_certs": self.ssl_ca_certs,
+        }
+
+
+class MemoryDBIndexConfig(BaseModel, DBCaseConfig):
+    metric_type: MetricType | None = None
+    insert_batch_size: int | None = None
+
+    def parse_metric(self) -> str:
+        if self.metric_type == MetricType.L2:
+            return "l2"
+        elif self.metric_type == MetricType.IP:
+            return "ip"
+        return "cosine"
+
+
+class MemoryDBHNSWConfig(MemoryDBIndexConfig):
+    M: int | None = 16
+    ef_construction: int | None = 64
+    ef_runtime: int | None = 10
+    index: IndexType = IndexType.HNSW
+
+    def index_param(self) -> dict:
+        return {
+            "metric": self.parse_metric(),
+            "index_type": self.index.value,
+            "m": self.M,
+            "ef_construction": self.ef_construction,
+        }
+
+    def search_param(self) -> dict:
+        return {
+            "ef_runtime": self.ef_runtime,
+        }

vectordb_bench-0.0.13/vectordb_bench/backend/clients/memorydb/memorydb.py

@@ -0,0 +1,254 @@
+import logging, time
+from contextlib import contextmanager
+from typing import Any, Generator, Optional, Tuple, Type
+from ..api import VectorDB, DBCaseConfig, IndexType
+from .config import MemoryDBIndexConfig
+import redis
+from redis import Redis
+from redis.cluster import RedisCluster
+from redis.commands.search.field import TagField, VectorField, NumericField
+from redis.commands.search.indexDefinition import IndexDefinition, IndexType
+from redis.commands.search.query import Query
+import numpy as np
+
+
+log = logging.getLogger(__name__)
+INDEX_NAME = "index"    # Vector Index Name
+
+class MemoryDB(VectorDB):
+    def __init__(
+            self,
+            dim: int,
+            db_config: dict,
+            db_case_config: MemoryDBIndexConfig,
+            drop_old: bool = False,
+            **kwargs
+        ):
+
+        self.db_config = db_config
+        self.case_config = db_case_config
+        self.collection_name = INDEX_NAME
+        self.target_nodes = RedisCluster.RANDOM if not self.db_config["cmd"] else None
+        self.insert_batch_size = db_case_config.insert_batch_size
+        self.dbsize = kwargs.get("num_rows")
+
+        # Create a MemoryDB connection, if db has password configured, add it to the connection here and in init():
+        log.info(f"Establishing connection to: {self.db_config}")
+        conn = self.get_client(primary=True)
+        log.info(f"Connection established: {conn}")
+        log.info(conn.execute_command("INFO server"))
+
+        if drop_old:
+            try:
+                log.info(f"MemoryDB client getting info for: {INDEX_NAME}")
+                info = conn.ft(INDEX_NAME).info()
+                log.info(f"Index info: {info}")
+            except redis.exceptions.ResponseError as e:
+                log.error(e)
+                drop_old = False
+                log.info(f"MemoryDB client drop_old collection: {self.collection_name}")
+            
+            log.info("Executing FLUSHALL")
+            conn.flushall()
+
+            # Since the default behaviour of FLUSHALL is asynchronous, wait for db to be empty
+            self.wait_until(self.wait_for_empty_db, 3, "", conn)
+            if not self.db_config["cmd"]:
+                replica_clients = self.get_client(replicas=True)
+                for rc, host in replica_clients:
+                    self.wait_until(self.wait_for_empty_db, 3, "", rc)
+                    log.debug(f"Flushall done in the host: {host}")
+                    rc.close()
+        
+        self.make_index(dim, conn)
+        conn.close()
+        conn = None
+
+    def make_index(self, vector_dimensions: int, conn: redis.Redis):
+        try:
+            # check to see if index exists
+            conn.ft(INDEX_NAME).info()
+        except Exception as e:
+            log.warn(f"Error getting info for index '{INDEX_NAME}': {e}")
+            index_param = self.case_config.index_param()
+            search_param = self.case_config.search_param()
+            vector_parameters = {  # Vector Index Type: FLAT or HNSW
+                "TYPE": "FLOAT32",
+                "DIM": vector_dimensions,  # Number of Vector Dimensions
+                "DISTANCE_METRIC": index_param["metric"],  # Vector Search Distance Metric
+            }
+            if index_param["m"]:
+                vector_parameters["M"] = index_param["m"]
+            if index_param["ef_construction"]:
+                vector_parameters["EF_CONSTRUCTION"] = index_param["ef_construction"]
+            if search_param["ef_runtime"]:
+                vector_parameters["EF_RUNTIME"] = search_param["ef_runtime"]
+
+            schema = (
+                TagField("id"),                   
+                NumericField("metadata"),              
+                VectorField("vector",   # Vector Field Name
+                    "HNSW", vector_parameters
+                ),
+            )
+
+            definition = IndexDefinition(index_type=IndexType.HASH)
+            rs = conn.ft(INDEX_NAME)
+            rs.create_index(schema, definition=definition)
+    
+    def get_client(self, **kwargs):
+        """
+        Gets either cluster connection or normal connection based on `cmd` flag.
+        CMD stands for Cluster Mode Disabled and is a "mode".
+        """
+        if not self.db_config["cmd"]:
+            # Cluster mode enabled
+
+            client = RedisCluster(
+                host=self.db_config["host"],
+                port=self.db_config["port"],
+                ssl=self.db_config["ssl"],
+                password=self.db_config["password"],
+                ssl_ca_certs=self.db_config["ssl_ca_certs"],
+                ssl_cert_reqs=None,
+            )
+
+            # Ping all nodes to create a connection
+            client.execute_command("PING", target_nodes=RedisCluster.ALL_NODES)
+            replicas = client.get_replicas()
+
+            if len(replicas) > 0:
+                # FT.SEARCH is a keyless command, use READONLY for replica connections
+                client.execute_command("READONLY", target_nodes=RedisCluster.REPLICAS)
+
+            if kwargs.get("primary", False):
+                client = client.get_primaries()[0].redis_connection
+
+            if kwargs.get("replicas", False):
+                # Return client and host name for each replica
+                return [(c.redis_connection, c.host) for c in replicas]
+
+        else:
+            client = Redis(
+                host=self.db_config["host"],
+                port=self.db_config["port"],
+                db=0,
+                ssl=self.db_config["ssl"],
+                password=self.db_config["password"],
+                ssl_ca_certs=self.db_config["ssl_ca_certs"],
+                ssl_cert_reqs=None,
+            )
+            client.execute_command("PING")
+        return client
+
+    @contextmanager
+    def init(self) -> Generator[None, None, None]:
+        """ create and destory connections to database.
+
+        Examples:
+            >>> with self.init():
+            >>>     self.insert_embeddings()
+        """
+        self.conn = self.get_client()
+        search_param = self.case_config.search_param()
+        if search_param["ef_runtime"]:
+            self.ef_runtime_str = f'EF_RUNTIME {search_param["ef_runtime"]}'
+        else:
+            self.ef_runtime_str = ""
+        yield
+        self.conn.close()
+        self.conn = None
+
+    def ready_to_load(self) -> bool:
+        pass
+
+    def optimize(self) -> None:
+        self._post_insert()
+
+    def insert_embeddings(
+        self,
+        embeddings: list[list[float]],
+        metadata: list[int],
+        **kwargs: Any,
+    ) -> Tuple[int, Optional[Exception]]:
+        """Insert embeddings into the database.
+        Should call self.init() first.
+        """
+
+        try:
+            with self.conn.pipeline(transaction=False) as pipe:
+                for i, embedding in enumerate(embeddings):
+                    embedding = np.array(embedding).astype(np.float32)
+                    pipe.hset(metadata[i], mapping = {
+                        "id": str(metadata[i]),
+                        "metadata": metadata[i], 
+                        "vector": embedding.tobytes(),
+                    })
+                    # Execute the pipe so we don't keep too much in memory at once
+                    if (i + 1) % self.insert_batch_size == 0:
+                        pipe.execute()
+
+                pipe.execute()
+                result_len = i + 1
+        except Exception as e:
+            return 0, e
+        
+        return result_len, None
+    
+    def _post_insert(self):
+        """Wait for indexing to finish"""
+        client = self.get_client(primary=True)
+        log.info("Waiting for background indexing to finish")
+        args = (self.wait_for_no_activity, 5, "", client)
+        self.wait_until(*args)
+        if not self.db_config["cmd"]:
+            replica_clients = self.get_client(replicas=True)
+            for rc, host_name in replica_clients:
+                args = (self.wait_for_no_activity, 5, "", rc)
+                self.wait_until(*args)
+                log.debug(f"Background indexing completed in the host: {host_name}")
+                rc.close()
+    
+    def wait_until(
+        self, condition, interval=5, message="Operation took too long", *args
+    ):
+        while not condition(*args):
+            time.sleep(interval)
+    
+    def wait_for_no_activity(self, client: redis.RedisCluster | redis.Redis):
+        return (
+            client.info("search")["search_background_indexing_status"] == "NO_ACTIVITY"
+        )
+    
+    def wait_for_empty_db(self, client: redis.RedisCluster | redis.Redis):
+        return client.execute_command("DBSIZE") == 0
+    
+    def search_embedding(
+        self,
+        query: list[float],
+        k: int = 10,
+        filters: dict | None = None,
+        timeout: int | None = None,
+        **kwargs: Any,
+    ) -> (list[int]):
+        assert self.conn is not None
+        
+        query_vector = np.array(query).astype(np.float32).tobytes()
+        query_obj = Query(f"*=>[KNN {k} @vector $vec]").return_fields("id").paging(0, k)
+        query_params = {"vec": query_vector}
+        
+        if filters:
+            # benchmark test filters of format: {'metadata': '>=10000', 'id': 10000}
+            # gets exact match for id, and range for metadata if they exist in filters
+            id_value = filters.get("id")
+            # Removing '>=' from the id_value: '>=10000'
+            metadata_value = filters.get("metadata")[2:]
+            if id_value and metadata_value:
+                query_obj = Query(f"(@metadata:[{metadata_value} +inf] @id:{ {id_value} })=>[KNN {k} @vector $vec]").return_fields("id").paging(0, k)
+            elif id_value:
+                #gets exact match for id
+                query_obj = Query(f"@id:{ {id_value} }=>[KNN {k} @vector $vec]").return_fields("id").paging(0, k)
+            else: #metadata only case, greater than or equal to metadata value
+                query_obj = Query(f"@metadata:[{metadata_value} +inf]=>[KNN {k} @vector $vec]").return_fields("id").paging(0, k)
+        res = self.conn.ft(INDEX_NAME).search(query_obj, query_params)
+        return [int(doc["id"]) for doc in res.docs]