memory/docker-compose.yml at development · dev-pratap-singh/memory · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
services:
  # PostgreSQL with pgvector extension
  postgres:
    image: ankane/pgvector:latest
    container_name: memory_postgres
    environment:
      POSTGRES_USER: ${DB_USER:-postgres}
      POSTGRES_PASSWORD: ${DB_PASSWORD:-postgres}
      POSTGRES_DB: ${DB_NAME:-memory_db}
      PGDATA: /var/lib/postgresql/data/pgdata
    ports:
      - "${DB_PORT:-5432}:5432"
    volumes:
      - postgres_data:/var/lib/postgresql/data
      - ./src/database/init_db.sql:/docker-entrypoint-initdb.d/init_db.sql
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U ${DB_USER:-postgres}"]
      interval: 10s
      timeout: 5s
      retries: 5
    networks:
      - memory_network

  # Redis for task queue and caching
  redis:
    image: redis:7-alpine
    container_name: memory_redis
    ports:
      - "${REDIS_PORT:-6379}:6379"
    volumes:
      - redis_data:/data
    command: redis-server --appendonly yes ${REDIS_PASSWORD:+--requirepass $REDIS_PASSWORD}
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 10s
      timeout: 5s
      retries: 5
    networks:
      - memory_network

  # Main application
  # For Linux with NVIDIA GPU: change dockerfile to Dockerfile.cuda
  app:
    build:
      context: .
      dockerfile: ${DOCKERFILE:-Dockerfile}  # Use Dockerfile.cuda for GPU on Linux
      target: ${BUILD_TARGET:-production}
    container_name: memory_app
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_NAME=${DB_NAME:-memory_db}
      - DB_USER=${DB_USER:-postgres}
      - DB_PASSWORD=${DB_PASSWORD:-postgres}
      - REDIS_HOST=redis
      - REDIS_PORT=6379
      - REDIS_PASSWORD=${REDIS_PASSWORD:-}
      - OPENAI_API_KEY=${OPENAI_API_KEY}
      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY:-}
      - USE_GPU=${USE_GPU:-false}  # Set to true if using Dockerfile.cuda
      - TRANSFORMERS_CACHE=/app/models/cache
      - ENVIRONMENT=${ENVIRONMENT:-production}
      - DEBUG=${DEBUG:-false}
    ports:
      - "${API_PORT:-8000}:8000"
      - "${METRICS_PORT:-9090}:9090"
    volumes:
      # Persistent data
      - ./data/conversations:/app/data/conversations
      - ./data/models:/app/data/models
      - ./data/backups:/app/data/backups
      - ./data/embeddings_cache:/app/data/embeddings_cache
      - ./logs:/app/logs
      # Model cache (can be large)
      - model_cache:/app/models/cache
      # Config (for development)
      - ./config.yaml:/app/config.yaml:ro
    networks:
      - memory_network
    # Uncomment below for NVIDIA GPU support (Linux only, use Dockerfile.cuda)
    # deploy:
    #   resources:
    #     reservations:
    #       devices:
    #         - driver: nvidia
    #           count: ${GPU_COUNT:-1}
    #           capabilities: [gpu]
    restart: unless-stopped

  # Celery worker for background tasks (training, etc.)
  worker:
    build:
      context: .
      dockerfile: ${DOCKERFILE:-Dockerfile}  # Use Dockerfile.cuda for GPU on Linux
      target: ${BUILD_TARGET:-production}
    container_name: memory_worker
    command: celery -A src.tasks.celery_app worker --loglevel=info --concurrency=${WORKER_CONCURRENCY:-2}
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    environment:
      - DB_HOST=postgres
      - DB_PORT=5432
      - DB_NAME=${DB_NAME:-memory_db}
      - DB_USER=${DB_USER:-postgres}
      - DB_PASSWORD=${DB_PASSWORD:-postgres}
      - REDIS_HOST=redis
      - REDIS_PORT=6379
      - REDIS_PASSWORD=${REDIS_PASSWORD:-}
      - USE_GPU=${USE_GPU:-false}  # Set to true if using Dockerfile.cuda
      - TRANSFORMERS_CACHE=/app/models/cache
    volumes:
      - ./data/conversations:/app/data/conversations
      - ./data/models:/app/data/models
      - ./data/backups:/app/data/backups
      - ./logs:/app/logs
      - model_cache:/app/models/cache
      - ./config.yaml:/app/config.yaml:ro
    networks:
      - memory_network
    # Uncomment below for NVIDIA GPU support (Linux only, use Dockerfile.cuda)
    # deploy:
    #   resources:
    #     reservations:
    #       devices:
    #         - driver: nvidia
    #           count: ${GPU_COUNT:-1}
    #           capabilities: [gpu]
    restart: unless-stopped

  # Celery beat for scheduled tasks
  beat:
    build:
      context: .
      dockerfile: ${DOCKERFILE:-Dockerfile}  # Use Dockerfile.cuda for GPU on Linux
      target: ${BUILD_TARGET:-production}
    container_name: memory_beat
    command: celery -A src.tasks.celery_app beat --loglevel=info
    depends_on:
      - redis
    environment:
      - REDIS_HOST=redis
      - REDIS_PORT=6379
      - REDIS_PASSWORD=${REDIS_PASSWORD:-}
    volumes:
      - ./config.yaml:/app/config.yaml:ro
    networks:
      - memory_network
    restart: unless-stopped

volumes:
  postgres_data:
    driver: local
  redis_data:
    driver: local
  model_cache:
    driver: local

networks:
  memory_network:
    driver: bridge