Productify

Appearance

Autoscaler Deployment ​

Complete deployment guide for the Productify Autoscaler components.

Components ​

The Autoscaler consists of two components:

  • Nomadscaler Plugin - Go-based Nomad Autoscaler plugin
  • Optimizer Service - Python-based MILP optimization service

Optimizer Deployment ​

Docker ​

bash
docker run -d \
  --name optimizer \
  -p 8000:8000 \
  -e CACHE_SIZE=10 \
  -e FORECAST_HORIZON=60 \
  ghcr.io/productifyfw/optimizer:latest

Docker Compose ​

yaml
version: "3.8"

services:
  optimizer:
    image: ghcr.io/productifyfw/optimizer:latest
    ports:
      - "8000:8000"
      - "9090:9090" # Prometheus metrics
    environment:
      CACHE_SIZE: "10"
      FORECAST_HORIZON: "60"
      LOG_LEVEL: "INFO"
    volumes:
      - ./config.ini:/app/config.ini:ro
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8000/health"]
      interval: 30s
      timeout: 10s
      retries: 3

Nomad ​

hcl
job "optimizer" {
  datacenters = ["dc1"]
  type = "service"

  group "optimizer" {
    count = 2  # For high availability

    network {
      port "http" {
        to = 8000
      }
      port "metrics" {
        to = 9090
      }
    }

    service {
      name = "optimizer"
      port = "http"

      tags = ["productify", "autoscaler"]

      check {
        type = "http"
        path = "/health"
        interval = "10s"
        timeout = "2s"
      }
    }

    task "server" {
      driver = "docker"

      config {
        image = "ghcr.io/productifyfw/optimizer:latest"
        ports = ["http", "metrics"]
      }

      env {
        CACHE_SIZE = "10"
        FORECAST_HORIZON = "60"
        LOG_LEVEL = "INFO"
      }

      resources {
        cpu    = 1000
        memory = 1024
      }
    }
  }
}

Future Support

Kubernetes deployment manifests will be added in a future release.

Nomadscaler Plugin Deployment ​

Installation ​

bash
# Build plugin
cd nomadscaler
./build.sh

# Install to Nomad Autoscaler
sudo cp ./bin/nomadscaler /opt/nomad-autoscaler/plugins/
sudo chmod +x /opt/nomad-autoscaler/plugins/nomadscaler

Nomad Autoscaler Configuration ​

hcl
# autoscaler.hcl
plugin_dir = "/opt/nomad-autoscaler/plugins"

nomad {
  address = "http://localhost:4646"
}

policy_dir = "/etc/nomad-autoscaler/policies"

http {
  bind_address = "0.0.0.0"
  bind_port    = 8080
}

log {
  level = "INFO"
}

Scaling Policy ​

hcl
# /etc/nomad-autoscaler/policies/web-app.hcl
scaling "horizontal" {
  enabled = true
  min     = 2
  max     = 20

  policy {
    evaluation_interval = "10s"
    cooldown           = "30s"

    target "nomadscaler" {
      namespace     = "production"
      job           = "web-app"
      group         = "frontend"
      optimizer_url = "http://optimizer:8000"
      cache_size    = 10
    }

    check "cpu" {
      source = "nomad-apm"
      query  = "avg_cpu_percent"

      strategy "target-value" {
        target = 75
      }
    }
  }
}

Running Nomad Autoscaler ​

bash
nomad-autoscaler agent \
  -config=/etc/nomad-autoscaler/autoscaler.hcl \
  -plugin-dir=/opt/nomad-autoscaler/plugins

Complete Stack Deployment ​

Docker Compose ​

yaml
version: "3.8"

services:
  optimizer:
    image: ghcr.io/productifyfw/optimizer:latest
    ports:
      - "8000:8000"
    environment:
      CACHE_SIZE: "10"
      FORECAST_HORIZON: "60"
    restart: unless-stopped

  nomad-autoscaler:
    image: hashicorp/nomad-autoscaler:latest
    volumes:
      - ./autoscaler.hcl:/etc/autoscaler.hcl:ro
      - ./policies:/policies:ro
      - ./plugins:/plugins:ro
    command: agent -config=/etc/autoscaler.hcl
    environment:
      NOMAD_ADDR: http://nomad:4646
    depends_on:
      - optimizer
    restart: unless-stopped

Nomad Job (Complete) ​

hcl
job "autoscaler-stack" {
  datacenters = ["dc1"]

  group "optimizer" {
    count = 2

    network {
      port "http" { to = 8000 }
      port "metrics" { to = 9090 }
    }

    service {
      name = "optimizer"
      port = "http"
      check {
        type = "http"
        path = "/health"
        interval = "10s"
      }
    }

    task "server" {
      driver = "docker"
      config {
        image = "ghcr.io/productifyfw/optimizer:latest"
        ports = ["http", "metrics"]
      }
      env {
        CACHE_SIZE = "10"
        FORECAST_HORIZON = "60"
      }
      resources {
        cpu    = 200
        memory = 256
      }
    }
  }

  group "autoscaler" {
    count = 1

    network {
      port "http" { to = 8080 }
    }

    task "agent" {
      driver = "docker"
      config {
        image = "hashicorp/nomad-autoscaler:latest"
        volumes = [
          "local/config.hcl:/etc/autoscaler.hcl",
          "local/policies:/policies",
          "local/plugins:/plugins"
        ]
        args = ["agent", "-config=/etc/autoscaler.hcl"]
      }

      template {
        data = file("autoscaler.hcl")
        destination = "local/config.hcl"
      }

      resources {
        cpu    = 200
        memory = 256
      }
    }
  }
}

Configuration ​

Optimizer Configuration ​

See Optimizer Configuration for details.

Key settings:

  • cache_size - Prediction cache size (default: 10)
  • forecast_horizon - Forecast window (default: 60)
  • sarimax_order - Time-series model parameters

Plugin Configuration ​

See Scaling Policies for details.

Key settings:

  • optimizer_url - Optimizer service URL
  • cache_size - Local cache size
  • evaluation_interval - How often to evaluate
  • cooldown - Minimum time between scaling actions

Health Checks ​

Optimizer Health ​

bash
curl http://localhost:8000/health

Nomad Autoscaler Status ​

bash
curl http://localhost:8080/v1/health

Plugin Status ​

bash
nomad-autoscaler policy list

Monitoring ​

Optimizer Metrics ​

Prometheus metrics available at /metrics:

optimizer_predictions_total
optimizer_forecast_latency_seconds
optimizer_optimization_latency_seconds
optimizer_errors_total

Nomad Autoscaler Metrics ​

nomad_autoscaler_policy_evaluation_total
nomad_autoscaler_scaling_actions_total
nomad_autoscaler_policy_evaluation_duration_seconds

Troubleshooting ​

Optimizer Not Responding ​

Check:

  • Service is running
  • Port 8000 is accessible
  • Health endpoint returns 200
  • Logs for errors

Plugin Not Loading ​

Verify:

  • Plugin binary in correct directory
  • Binary has execute permissions
  • Plugin version matches autoscaler version
  • Logs show plugin registration

No Scaling Actions ​

Debug:

  • Policy is enabled
  • Metrics are being collected
  • Optimizer is reachable
  • Cooldown period hasn't been triggered
  • Min/max bounds allow scaling

See Also ​