Refactor docker compose of the dataprep microservice

comps/dataprep/deployment/docker_compose/compose.yaml

@@ -0,0 +1,258 @@
+# Copyright (C) 2024 Intel Corporation
+# SPDX-License-Identifier: Apache-2.0
+
+
+include:
+  - ../../../third_parties/elasticsearch/deployment/docker_compose/compose.yaml
+  - ../../../third_parties/neo4j/deployment/docker_compose/compose.yaml
+  - ../../../third_parties/opensearch/deployment/docker_compose/compose.yaml
+  - ../../../third_parties/pgvector/deployment/docker_compose/compose.yaml
+  - ../../../third_parties/qdrant/deployment/docker_compose/compose.yaml
+  - ../../../third_parties/redis/deployment/docker_compose/compose.yaml
+  - ../../../third_parties/vdms/deployment/docker_compose/compose.yaml
+  - ../../../third_parties/tgi/deployment/docker_compose/compose.yaml
+  - ../../../third_parties/tei/deployment/docker_compose/compose.yaml
+
+services:
+
+  dataprep-elasticsearch:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-elasticsearch
+    ports:
+      - "${DATAPREP_PORT:-11100}:5000"
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_ELASTICSEARCH"
+      ES_CONNECTION_STRING: ${ES_CONNECTION_STRING}
+      INDEX_NAME: ${INDEX_NAME}
+      TEI_ENDPOINT: ${TEI_ENDPOINT}
+      HUGGINGFACEHUB_API_TOKEN: ${HF_TOKEN}
+    restart: unless-stopped
+    depends_on:
+      elasticsearch-vector-db:
+        condition: service_healthy
+
+  dataprep-milvus:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-milvus-server
+    ports:
+      - "${DATAPREP_PORT:-11101}:5000"
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_MILVUS"
+      MILVUS_HOST: ${MILVUS_HOST}
+      TEI_EMBEDDING_ENDPOINT: ${TEI_EMBEDDING_ENDPOINT}
+      HUGGINGFACEHUB_API_TOKEN: ${HF_TOKEN}
+    restart: unless-stopped
+    depends_on:
+      tei-embedding-serving:
+        condition: service_healthy
+
+  dataprep-neo4j-llamaindex:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-neo4j-llamaindex
+    depends_on:
+      neo4j-apoc:
+        condition: service_healthy
+      tgi-gaudi-server:
+        condition: service_healthy
+      tei-embedding-serving:
+        condition: service_healthy
+    ports:
+      - "${DATAPREP_PORT:-11103}:5000"
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      host_ip: ${host_ip}
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_NEO4J_LLAMAINDEX"
+      NEO4J_URL: ${NEO4J_URL}
+      NEO4J_USERNAME: ${NEO4J_USERNAME}
+      NEO4J_PASSWORD: ${NEO4J_PASSWORD}
+      TGI_LLM_ENDPOINT: ${TGI_LLM_ENDPOINT}
+      TEI_EMBEDDING_ENDPOINT: ${TEI_EMBEDDING_ENDPOINT}
+      OPENAI_API_KEY: ${OPENAI_API_KEY}
+      OPENAI_EMBEDDING_MODEL: ${OPENAI_EMBEDDING_MODEL}
+      OPENAI_LLM_MODEL: ${OPENAI_LLM_MODEL}
+      EMBEDDING_MODEL_ID: ${EMBEDDING_MODEL_ID}
+      LLM_MODEL_ID: ${LLM_MODEL_ID}
+      LOGFLAG: ${LOGFLAG}
+      HUGGINGFACEHUB_API_TOKEN: ${HF_TOKEN}
+    restart: unless-stopped
+
+  dataprep-opensearch:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-opensearch-server
+    ports:
+      - "${DATAPREP_PORT:-11104}:5000"
+    depends_on:
+      opensearch-vector-db:
+        condition: service_healthy
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      OPENSEARCH_INITIAL_ADMIN_PASSWORD: ${OPENSEARCH_INITIAL_ADMIN_PASSWORD}
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_OPENSEARCH"
+      OPENSEARCH_URL: ${OPENSEARCH_URL}
+      INDEX_NAME: ${INDEX_NAME}
+      HUGGINGFACEHUB_API_TOKEN: ${HF_TOKEN}
+    restart: unless-stopped
+    security_opt:
+      - no-new-privileges:true
+
+  dataprep-pgvector:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-pgvector-server
+    ports:
+      - "${DATAPREP_PORT:-11105}:5000"
+    depends_on:
+      pgvector-db:
+        condition: service_healthy
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_PGVECTOR"
+      PG_CONNECTION_STRING: ${PG_CONNECTION_STRING}
+    restart: unless-stopped
+
+  dataprep-pinecone:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-pinecone-server
+    ports:
+      - "${DATAPREP_PORT:-11106}:5000"
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_PINECONE"
+      PINECONE_API_KEY: ${PINECONE_API_KEY}
+      PINECONE_INDEX_NAME: ${PINECONE_INDEX_NAME}
+      HUGGINGFACEHUB_API_TOKEN: ${HF_TOKEN}
+    restart: unless-stopped
+
+  dataprep-qdrant:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-qdrant-server
+    depends_on:
+      qdrant-vector-db:
+        condition: service_healthy
+      tei-embedding-serving:
+        condition: service_healthy
+    ports:
+      - "${DATAPREP_PORT:-11107}:5000"
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_QDRANT"
+      QDRANT_HOST: ${QDRANT_HOST}
+      QDRANT_PORT: ${QDRANT_PORT}
+      COLLECTION_NAME: ${COLLECTION_NAME}
+      TEI_ENDPOINT: ${TEI_EMBEDDING_ENDPOINT}
+      HUGGINGFACEHUB_API_TOKEN: ${HF_TOKEN}
+    restart: unless-stopped
+
+  dataprep-redis:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-redis-server
+    depends_on:
+      redis-vector-db:
+        condition: service_healthy
+      tei-embedding-serving:
+        condition: service_healthy
+    ports:
+      - "${DATAPREP_PORT:-11108}:5000"
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_REDIS"
+      REDIS_HOST: ${REDIS_HOST}
+      REDIS_PORT: ${REDIS_PORT}
+      REDIS_URL: ${REDIS_URL}
+      TEI_EMBEDDING_ENDPOINT: ${TEI_EMBEDDING_ENDPOINT}
+      INDEX_NAME: ${INDEX_NAME}
+      HUGGINGFACEHUB_API_TOKEN: ${HF_TOKEN}
+    restart: unless-stopped
+
+  dataprep-multimodal-redis:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-multimodal-redis-server
+    ports:
+      - "${DATAPREP_PORT:-11109}:5000"
+    depends_on:
+      redis-vector-db:
+        condition: service_healthy
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      MULTIMODAL_DATAPREP: true
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_MULTIMODALREDIS"
+      REDIS_HOST: ${REDIS_HOST}
+      REDIS_PORT: ${REDIS_PORT}
+      REDIS_URL: ${REDIS_URL}
+      INDEX_NAME: ${INDEX_NAME}
+      LVM_ENDPOINT: ${LVM_ENDPOINT}
+      HUGGINGFACEHUB_API_TOKEN: ${HF_TOKEN}
+    restart: unless-stopped
+
+  dataprep-vdms:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-vdms-server
+    ports:
+      - "${DATAPREP_PORT:-11110}:5000"
+    depends_on:
+      vdms-vector-db:
+        condition: service_healthy
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_VDMS"
+      VDMS_HOST: ${VDMS_HOST}
+      VDMS_PORT: ${VDMS_PORT}
+      COLLECTION_NAME: ${COLLECTION_NAME}
+      HUGGINGFACEHUB_API_TOKEN: ${HF_TOKEN}
+    restart: unless-stopped
+
+  dataprep-vdms-multimodal:
+    image: ${REGISTRY:-opea}/dataprep:${TAG:-latest}
+    container_name: dataprep-vdms-multimodal-server
+    ports:
+      - "${DATAPREP_PORT:-11111}:5000"
+    depends_on:
+      vdms-vector-db:
+        condition: service_healthy
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      MULTIMODAL_DATAPREP: true
+      DATAPREP_COMPONENT_NAME: "OPEA_DATAPREP_MULTIMODALVDMS"
+      VDMS_HOST: ${VDMS_HOST}
+      VDMS_PORT: ${VDMS_PORT}
+      COLLECTION_NAME: ${INDEX_NAME}
+    restart: unless-stopped
+
+networks:
+  default:
+    driver: bridge
+  opensearch-net: