VLM Installation

This page informs how to deploy the docker compose of snapshot platform as the VisionAIre Stream analytics requirements for every analytics run by VLM deployment.

How to Setup the Deployment?

Create a dedicated folder for this installation to help organize your deployment.

1. Configuration of docker-compose.yml

version: '3.3'
services:
  node1:
    image: "${VLM_IMAGE}"
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]
    pid: host
    network_mode: host
    cap_add:
      - SYS_PTRACE
    command: [
      httpserver,
      --listen-port, "${FREMIS1_LISTEN_PORT}",
      --listen-port-monitoring, "${FREMIS1_LISTEN_PORT_MONITORING}",
      --verbose,
    ]
    healthcheck:
      test: ["CMD", "curl", "-f", "http://0.0.0.0:${FREMIS1_LISTEN_PORT}/healthcheck"]
      interval: 5s
      timeout: 3s
      retries: 20
  coordinator:
    image: "${VLM_IMAGE}"
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]
    pid: host
    network_mode: host
    cap_add:
      - SYS_PTRACE
    command: [
      coordinator,
      --listen-port, "${COORDINATOR_LISTEN_PORT}",
      --listen-port-monitoring, "${COORDINATOR_LISTEN_PORT_MONITORING}",
      --config-path, "/etc/nodeflux/config.yml",
      --verbose,
    ]
    volumes:
      - ${PWD}/config.yml:/etc/nodeflux/config.yml
    depends_on:
      node1:
        condition: service_healthy

2. Settings of config.yml

version: "v1"
nodes:
- address: "0.0.0.0:5051"
  analytic_id: "NFFS-VLM"

3. Keep the analytics image and port addresses into .env file

export VLM_IMAGE=registry.gitlab.com/nodefluxio/cloud/analytics/pipelines/vlm-pipeline:on-premise-0.1.0 ## for float16
export VLM_IMAGE=registry.gitlab.com/nodefluxio/cloud/analytics/pipelines/vlm-pipeline:on-premise-0.1.0-full ## for full precision

export COORDINATOR_LISTEN_PORT=4013
export COORDINATOR_LISTEN_PORT_MONITORING=5013

export FREMIS1_LISTEN_PORT=5051
export FREMIS1_LISTEN_PORT_MONITORING=6061

Use the image vlm-pipeline:on-premise-0.1.0 if you are using VLM to run basic analytics

Use the image vlm-pipeline:on-premise-0.1.0-full if you are using VLM to run advanced analytics

4. Run Image

docker-compose up -d --build