fAIr/docker-compose.dev.yml at develop · Spatialnode/fAIr · 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
# fair-dev instance override, layered on top of docker-compose.yml.
services:
  frontend-assets:
    environment:
      VITE_BASE_API_URL: ${VITE_BASE_API_URL}
      VITE_AUTH_PROVIDER: ${VITE_AUTH_PROVIDER}
      VITE_HANKO_URL: ${VITE_HANKO_URL}
      VITE_FAIR_STAC_CATALOG_BASE_URL: ${VITE_FAIR_STAC_CATALOG_BASE_URL}
      VITE_NODE_ENV: ${VITE_NODE_ENV}
      VITE_FAIR_PROD_URL: ${VITE_FAIR_PROD_URL}

  caddy:
    image: caddy:2-alpine
    restart: unless-stopped
    ports:
      - "80:80"
      - "443:443"
    extra_hosts:
      - "host.docker.internal:host-gateway"
    configs:
      - source: caddyfile
        target: /etc/caddy/Caddyfile
    volumes:
      - caddy_data:/data
      - caddy_config:/config
    depends_on:
      - api

  postgres:
    restart: unless-stopped

  minio:
    restart: unless-stopped

  stac:
    restart: unless-stopped

  mlflow:
    restart: unless-stopped

  zenml:
    restart: unless-stopped

  api:
    restart: unless-stopped
    entrypoint:
      [
        "sh",
        "-c",
        'zenml init && zenml stack set fair-compose && python manage.py collectstatic --noinput && exec "$$@"',
        "--",
      ]
    command:
      - gunicorn
      - --bind=0.0.0.0:8000
      - --workers=3
      - --timeout=120
      - config.wsgi:application

  worker:
    restart: unless-stopped

configs:
  # ZenML artifact store on real AWS (same bucket + instance IAM role as backend storage).
  zenml_stack:
    content: |
      stack_name: fair-compose
      zenml_version: 0.94.2
      components:
        orchestrator:
          name: docker_compose
          flavor: local_docker
          type: orchestrator
          configuration:
            synchronous: false
            run_args:
              network: fair-net
              user: "0"
              environment:
                NVIDIA_VISIBLE_DEVICES: all
                NVIDIA_DRIVER_CAPABILITIES: compute,utility
        artifact_store:
          name: s3_compose
          flavor: s3
          type: artifact_store
          configuration:
            path: s3://${BUCKET_NAME:-fair-dev}/${PARENT_BUCKET_FOLDER:-dev}/zenml
        experiment_tracker:
          name: mlflow_compose
          flavor: mlflow
          type: experiment_tracker
          configuration:
            tracking_uri: "http://mlflow:5000"
            tracking_username: admin
            tracking_password: admin

  caddyfile:
    content: |
      {
        email ${CADDY_ACME_EMAIL:-sysadmin@hotosm.org}
        on_demand_tls {
          ask http://127.0.0.1:5555/check
        }
      }
      ${PUBLIC_DOMAIN:-dev.ai.hotosm.org}, fair-dev.hotosm.org {
        handle_path /stac/* {
          reverse_proxy stac:8080
        }
        reverse_proxy api:8000
      }
      stac.${PUBLIC_DOMAIN:-dev.ai.hotosm.org} {
        reverse_proxy stac:8080
      }
      # Per-model knative predict endpoints. Kourier is HTTP-only (NodePort), so
      # Caddy terminates TLS per host on demand and proxies to the node ingress.
      :5555 {
        respond /check 200
      }
      *.predict.${PUBLIC_DOMAIN:-dev.ai.hotosm.org} {
        tls {
          on_demand
        }
        reverse_proxy ${KNATIVE_KOURIER_ADDR:-host.docker.internal:31751}
      }
      # mlflow.${PUBLIC_DOMAIN:-dev.ai.hotosm.org} { reverse_proxy mlflow:5000 }
      # zenml.${PUBLIC_DOMAIN:-dev.ai.hotosm.org} { reverse_proxy zenml:8080 }
      # minio.${PUBLIC_DOMAIN:-dev.ai.hotosm.org} { reverse_proxy minio:9000 }
      # console.${PUBLIC_DOMAIN:-dev.ai.hotosm.org} { reverse_proxy minio:9001 }

volumes:
  caddy_data:
  caddy_config: