feat: add Valkey as a vector store backend

.github/workflows/vdb-tests-full.yml

@@ -56,13 +56,14 @@ jobs:
 #            tidb
 #            tiflash
 
+
 #      - name: Check VDB Ready (TiDB)
 #        run: uv run --project api python api/providers/vdb/tidb-vector/tests/integration_tests/check_tiflash_ready.py
 
       - name: Test Vector Stores
         run: |
           uv run --project api pytest \
             --start-vdb \
-            --vdb-services "weaviate,qdrant,couchbase-server,etcd,minio,milvus-standalone,pgvecto-rs,pgvector,chroma,elasticsearch,oceanbase" \
+            --vdb-services "weaviate,qdrant,couchbase-server,etcd,minio,milvus-standalone,pgvecto-rs,pgvector,chroma,elasticsearch,oceanbase,valkey" \
             --timeout "${PYTEST_TIMEOUT:-180}" \
             api/providers/vdb/*/tests/integration_tests

.github/workflows/vdb-tests.yml

@@ -53,6 +53,7 @@ jobs:
 #            tidb
 #            tiflash
 
+
 #      - name: Check VDB Ready (TiDB)
 #        run: uv run --project api python api/providers/vdb/tidb-vector/tests/integration_tests/check_tiflash_ready.py
 
@@ -64,4 +65,5 @@ jobs:
             api/providers/vdb/vdb-chroma/tests/integration_tests \
             api/providers/vdb/vdb-pgvector/tests/integration_tests \
             api/providers/vdb/vdb-qdrant/tests/integration_tests \
+            api/providers/vdb/vdb-valkey/tests/integration_tests \
             api/providers/vdb/vdb-weaviate/tests/integration_tests

api/.env.example

@@ -201,7 +201,7 @@ CONSOLE_CORS_ALLOW_ORIGINS=http://localhost:3000,*
 COOKIE_DOMAIN=
 
 # Vector database configuration
-# Supported values are `weaviate`, `oceanbase`, `qdrant`, `milvus`, `myscale`, `relyt`, `pgvector`, `pgvecto-rs`, `chroma`, `opensearch`, `oracle`, `tencent`, `elasticsearch`, `elasticsearch-ja`, `analyticdb`, `couchbase`, `vikingdb`,  `opengauss`, `tablestore`,`vastbase`,`tidb`,`tidb_on_qdrant`,`baidu`,`lindorm`,`huawei_cloud`,`upstash`, `matrixone`, `hologres`.
+# Supported values are `weaviate`, `oceanbase`, `qdrant`, `milvus`, `myscale`, `relyt`, `pgvector`, `pgvecto-rs`, `chroma`, `opensearch`, `oracle`, `tencent`, `elasticsearch`, `elasticsearch-ja`, `analyticdb`, `couchbase`, `vikingdb`,  `opengauss`, `tablestore`,`vastbase`,`tidb`,`tidb_on_qdrant`,`baidu`,`lindorm`,`huawei_cloud`,`upstash`, `matrixone`, `hologres`, `valkey`.
 VECTOR_STORE=weaviate
 # Prefix used to create collection name in vector database
 VECTOR_INDEX_NAME_PREFIX=Vector_index
@@ -231,6 +231,14 @@ QDRANT_GRPC_ENABLED=false
 QDRANT_GRPC_PORT=6334
 QDRANT_REPLICATION_FACTOR=1
 
+# Valkey configuration, only available when VECTOR_STORE is `valkey`
+VALKEY_HOST=localhost
+VALKEY_PORT=6379
+VALKEY_PASSWORD=
+VALKEY_DB=0
+VALKEY_USE_SSL=false
+VALKEY_DISTANCE_METRIC=COSINE
+
 #Couchbase configuration
 COUCHBASE_CONNECTION_STRING=127.0.0.1
 COUCHBASE_USER=Administrator

api/configs/middleware/__init__.py

@@ -46,6 +46,7 @@
 from .vdb.tidb_on_qdrant_config import TidbOnQdrantConfig
 from .vdb.tidb_vector_config import TiDBVectorConfig
 from .vdb.upstash_config import UpstashConfig
+from .vdb.valkey_config import ValkeyConfig
 from .vdb.vastbase_vector_config import VastbaseVectorConfig
 from .vdb.vikingdb_config import VikingDBConfig
 from .vdb.weaviate_config import WeaviateConfig
@@ -407,5 +408,6 @@ class MiddlewareConfig(
     TableStoreConfig,
     DatasetQueueMonitorConfig,
     MatrixoneConfig,
+    ValkeyConfig,
 ):
     pass

api/configs/middleware/vdb/valkey_config.py

@@ -0,0 +1,38 @@
+from typing import Literal
+
+from pydantic import Field, NonNegativeInt, PositiveInt
+from pydantic_settings import BaseSettings
+
+
+class ValkeyConfig(BaseSettings):
+    """Configuration settings for Valkey vector database (valkey-search module)."""
+
+    VALKEY_HOST: str = Field(
+        description="Hostname or IP address of the Valkey server.",
+        default="localhost",
+    )
+
+    VALKEY_PORT: PositiveInt = Field(
+        description="Port number for the Valkey server (default is 6379).",
+        default=6379,
+    )
+
+    VALKEY_PASSWORD: str = Field(
+        description="Password for authenticating with the Valkey server.",
+        default="",
+    )
+
+    VALKEY_DB: NonNegativeInt = Field(
+        description="Valkey database number to use (default is 0).",
+        default=0,
+    )
+
+    VALKEY_USE_SSL: bool = Field(
+        description="Whether to use SSL/TLS for the Valkey connection.",
+        default=False,
+    )
+
+    VALKEY_DISTANCE_METRIC: Literal["COSINE", "L2", "IP"] = Field(
+        description="Distance metric for vector similarity search. Options: COSINE, L2, IP.",
+        default="COSINE",
+    )

api/core/rag/datasource/vdb/vector_type.py

@@ -33,5 +33,6 @@ class VectorType(StrEnum):
     HUAWEI_CLOUD = "huawei_cloud"
     MATRIXONE = "matrixone"
     CLICKZETTA = "clickzetta"
+    VALKEY = "valkey"
     IRIS = "iris"
     HOLOGRES = "hologres"

api/providers/vdb/vdb-valkey/pyproject.toml

@@ -0,0 +1,14 @@
+[project]
+name = "dify-vdb-valkey"
+version = "0.0.1"
+
+dependencies = [
+    "valkey-glide>=1.3.0",
+]
+description = "Dify vector store backend (dify-vdb-valkey)."
+
+[project.entry-points."dify.vector_backends"]
+valkey = "dify_vdb_valkey.valkey_vector:ValkeyVectorFactory"
+
+[tool.setuptools.packages.find]
+where = ["src"]

api/providers/vdb/vdb-valkey/src/dify_vdb_valkey/__init__.py

@@ -0,0 +1 @@
+