Judoscale

https://judoscale.com/docs
Admin
Judoscale is an intelligent autoscaling platform that optimizes dyno scaling for Heroku, AWS,...
Tokens:23,239
Snippets:333
Trust Score:8.4
Update:6 months ago
Show doc for...
Tokens:
Raw
### Full Process Type Configuration Example

Source: https://judoscale.com/docs/api-docs

An example showing all available settings for a web process type, including detailed autoscaling metrics.

```json
{
  "process_type": {
    "name": "web",
    "autoscaling_enabled": true,
    "range_min": 2,
    "range_max": 12,
    "upscale_threshold": 100,
    "upscale_quantity": 2,
    "upscale_interval_seconds": 120,
    "upscale_sensitivity_seconds": 10,
    "downscale_threshold": 50,
    "downscale_quantity": 1,
    "downscale_interval_seconds": 600,
    "autoscale_metrics": [
      {
        "metric": "adapter_queue_time",
        "upscale_threshold": 100,
        "downscale_threshold": 50
      }
    ]
  }
}
```

--------------------------------

### Example Utilization Calculation

Source: https://judoscale.com/docs/understanding-utilization

Illustrates how to calculate overall utilization based on individual instance utilizations. This helps in understanding how to set autoscaling thresholds.

```text
Instance 1: 3 processes handling requests, 75% utilization
Instance 2: 1 process handling requests,   25% utilization
Instance 3: 2 processes handling requests, 50% utilization
```

```text
Instance 1: 3 processes handling requests, 75% utilization
Instance 2: 3 processes handling requests, 75% utilization
# or
Instance 1: 4 processes handling requests, 100% utilization
Instance 2: 2 processes handling requests, 50% utilization
```

--------------------------------

### GET /settings

Source: https://judoscale.com/docs/api-docs

Retrieves a snapshot of your current app settings, including process type configurations and autoscaling metrics.

```APIDOC
## GET /settings

### Description
This endpoint returns a snapshot of your current app settings.

### Method
GET

### Endpoint
/settings

### Query Parameters
#### backward_compatible_attributes (boolean) - Optional - Excludes legacy attributes such as the top-level `upscale_threshold` and `downscale_threshold` keys. Use this flag when you plan to pipe the response directly into `PATCH /process_type`. The payload produced with this flag contains only the modern `autoscale_metrics` collection.

### Response
#### Success Response (200)
- **slack_webhook_url** (string) - The Slack webhook URL for notifications.
- **name** (string) - The name of the application.
- **process_types** (array) - A list of process type configurations.
  - **name** (string) - The name of the process type (e.g., "web", "worker").
  - **autoscaling_enabled** (boolean) - Whether autoscaling is enabled for this process type.
  - **current_scale** (integer) - The current number of running instances.
  - **range_min** (integer) - The minimum number of instances allowed.
  - **range_max** (integer) - The maximum number of instances allowed.
  - **upscale_threshold** (number) - The threshold for upscaling (may be deprecated in favor of `autoscale_metrics`).
  - **upscale_quantity** (integer) - The number of instances to add when upscaling.
  - **upscale_interval_seconds** (integer) - The interval in seconds between upscale attempts.
  - **upscale_sensitivity_seconds** (integer) - The sensitivity in seconds for upscaling.
  - **downscale_threshold** (number) - The threshold for downscaling (may be deprecated in favor of `autoscale_metrics`).
  - **downscale_quantity** (integer) - The number of instances to remove when downscaling.
  - **downscale_interval_seconds** (integer) - The interval in seconds between downscale attempts.
  - **autoscale_metrics** (array) - A list of autoscaling metrics and their thresholds.
    - **metric** (string) - The name of the metric (e.g., "adapter_queue_time", "adapter_utilization").
    - **upscale_threshold** (number) - The threshold for upscaling for this metric.
    - **downscale_threshold** (number) - The threshold for downscaling for this metric.
  - **all_queues** (boolean) - Indicates if all queues are monitored for worker processes.
  - **queue_names** (array) - A list of specific queue names to monitor for worker processes.

### Response Example
```json
{
  "slack_webhook_url": "",
  "name": "example",
  "process_types": [
    {
      "name": "web",
      "autoscaling_enabled": true,
      "current_scale": 3,
      "range_min": 2,
      "range_max": 12,
      "upscale_threshold": 100,
      "upscale_quantity": 2,
      "upscale_interval_seconds": 120,
      "upscale_sensitivity_seconds": 10,
      "downscale_threshold": 50,
      "downscale_quantity": 1,
      "downscale_interval_seconds": 600,
      "autoscale_metrics": [
        {
          "metric": "adapter_queue_time",
          "upscale_threshold": 100,
          "downscale_threshold": 50
        }
      ]
    },
    {
      "name": "worker",
      "autoscaling_enabled": true,
      "current_scale": 1,
      "all_queues": false,
      "queue_names": ["default", "high"],
      "range_min": 1,
      "range_max": 6,
      "upscale_threshold": 10000,
      "upscale_quantity": 2,
      "upscale_interval_seconds": 120,
      "upscale_sensitivity_seconds": 10,
      "downscale_threshold": 1000,
      "downscale_quantity": 1,
      "downscale_interval_seconds": 600,
      "autoscale_metrics": [
        {
          "metric": "adapter_queue_time",
          "upscale_threshold": 10000,
          "downscale_threshold": 1000
        }
      ]
    }
  ]
}
```
```

--------------------------------

### GET /services

Source: https://judoscale.com/docs/custom-platform-getting-started

Retrieves a list of all services across all applications and environments.

```APIDOC
## GET /services

### Description
Returns the list of all your services, across all your applications / environments.

### Method
GET

### Endpoint
`/services`

### Parameters
None

### Request Example
```
GET <api_url>/services
```

### Response
#### Success Response (200)
- **id** (string) - Unique identifier for the service.
- **group** (string) - Identifier to group services (e.g., application or environment name).
- **scale** (integer) - The current scale for the service.
- **web** (boolean) - Optional. Must be `true` to identify web services; can be omitted otherwise. Worker processes should omit this flag or set it to `false`.

#### Response Example
```json
[
  { "id": "web-stag", "group": "custom-app-staging", "scale": 2, "web": true },
  { "id": "worker-stag", "group": "custom-app-staging", "scale": 1 },
  { "id": "web-prod", "group": "custom-app-production", "scale": 5, "web": true },
  { "id": "worker-prod", "group": "custom-app-production", "scale": 2 }
]
```
```

--------------------------------

### List Services Endpoint Request

Source: https://judoscale.com/docs/custom-platform-getting-started

Sample GET request to retrieve the list of services.

```text
GET <api_url>/services
```

--------------------------------

### Process Type Configuration Response

Source: https://judoscale.com/docs/api-docs

Example JSON response when retrieving process type details. It shows all available autoscaling settings and metric configurations.

```json
{
  "process_type": {
    "name": "web",
    "autoscaling_enabled": true,
    "current_scale": 3,
    "range_min": 2,
    "range_max": 12,
    "upscale_quantity": 2,
    "upscale_interval_seconds": 120,
    "upscale_sensitivity_seconds": 10,
    "downscale_quantity": 1,
    "downscale_interval_seconds": 600,
    "autoscale_metrics": [
      {
        "metric": "adapter_queue_time",
        "upscale_threshold": 100,
        "downscale_threshold": 50
      }
    ]
  }
}
```

--------------------------------

### Settings API Response Structure

Source: https://judoscale.com/docs/api-docs

Example JSON response for the settings endpoint showing process types and autoscale metrics.

```json
{
  "slack_webhook_url": "",
  "name": "example",
  "process_types": [
    {
      "name": "web",
      "autoscaling_enabled": true,
      "current_scale": 3,
      "range_min": 2,
      "range_max": 12,
      "upscale_threshold": 100,
      "upscale_quantity": 2,
      "upscale_interval_seconds": 120,
      "upscale_sensitivity_seconds": 10,
      "downscale_threshold": 50,
      "downscale_quantity": 1,
      "downscale_interval_seconds": 600,
      "autoscale_metrics": [
        {
          "metric": "adapter_queue_time",
          "upscale_threshold": 100,
          "downscale_threshold": 50
        }
      ]
    },
    {
      "name": "worker",
      "autoscaling_enabled": true,
      "current_scale": 1,
      "all_queues": false,
      "queue_names": ["default", "high"],
      "range_min": 1,
      "range_max": 6,
      "upscale_threshold": 10000,
      "upscale_quantity": 2,
      "upscale_interval_seconds": 120,
      "upscale_sensitivity_seconds": 10,
      "downscale_threshold": 1000,
      "downscale_quantity": 1,
      "downscale_interval_seconds": 600,
      "autoscale_metrics": [
        {
          "metric": "adapter_queue_time",
          "upscale_threshold": 10000,
          "downscale_threshold": 1000
        }
      ]
    }
  ]
}
```

```json
{
  "process_types": [
    {
      "name": "web",
      "autoscaling_enabled": true,
      "current_scale": 3,
      "range_min": 2,
      "range_max": 12,
      "upscale_quantity": 2,
      "upscale_interval_seconds": 120,
      "upscale_sensitivity_seconds": 10,
      "downscale_quantity": 1,
      "downscale_interval_seconds": 600,
      "autoscale_metrics": [
        {
          "metric": "adapter_queue_time",
          "upscale_threshold": 100,
          "downscale_threshold": 50
        },
        {
          "metric": "adapter_utilization",
          "upscale_threshold": 60,
          "downscale_threshold": 40
        }
      ]
    }
  ]
}
```

--------------------------------

### Create File System Queue Metrics Collector

Source: https://judoscale.com/docs/custom-worker-python

Implement a custom metrics collector for a file-based job queue. This example defines a `FileSystemQueueMetricsCollector` that calculates queue time based on the oldest job file's modification time in a specified directory structure. Ensure the `judoscale.core.metrics_collectors` and `judoscale.core.metrics` modules are imported.

```Python
from pathlib
from datetime
from judoscale.core.metrics_collectors
from judoscale.core.metrics

DEFAULTS = {
    "ENABLED": True,
    "MAX_QUEUES": 20,
    "QUEUES": [],
    "TRACK_BUSY_JOBS": False,
}

class FileSystemQueueMetricsCollector(JobMetricsCollector):
    def __init__(self, config: Config):
        super().__init__(config=config)
        # Override the defaults with any config values provided by the user
        self.config["FSQUEUE"] = {**DEFAULTS, **self.config.get("FSQUEUE", {})}

    @property
    def should_collect(self) -> bool:
        # A sensible default is provided in `JobMetricsCollector.should_collect`,
        # but you can override it here if needed.
        return True

    def collect(self) -> List[Metric]:
        metrics = []

        if not self.should_collect:
            return metrics

        for queue_path in Path("/tmp/queues").glob("*"):
            queue_name = queue_path.stem

            # Queue time is based on the oldest job in the queue
            job_enqueued_times = [
                path.stat().st_mtime for path in queue_path.glob("*")
            ]

            if not job_enqueued_times:
                continue

            oldest_job_ts = min(job_enqueued_times)

            # This is the important part!
            metrics.append(Metric.for_queue(queue_name, oldest_job_ts))

        # Also very important to return the metrics
        return metrics

```

--------------------------------

### Webhook Message with Dashboard Link

Source: https://judoscale.com/docs/slack-notifications

Example of a notification message containing a Slack-formatted link to the Judoscale dashboard.

```text
Judoscale is scaling UP - scaling to 4 <https://app.judoscale.com/apps/123-your-app-production|web dynos> for your-app-production
```

--------------------------------

### GET /process_type/:name

Source: https://judoscale.com/docs/api-docs

Retrieves the current autoscaling configuration for a specific process type.

```APIDOC
## GET /process_type/:name

### Description
Retrieves the current autoscaling configuration for a specific process type.

### Method
GET

### Endpoint
/process_type/:name

### Parameters
#### Path Parameters
- **name** (string) - Required - The name of the process type (e.g., "web", "worker").

### Response
#### Success Response (200)
- **process_type** (object) - The configuration object for the process type.

#### Response Example
{
  "process_type": {
    "name": "web",
    "autoscaling_enabled": true,
    "current_scale": 3,
    "range_min": 2,
    "range_max": 12,
    "upscale_quantity": 2,
    "upscale_interval_seconds": 120,
    "upscale_sensitivity_seconds": 10,
    "downscale_quantity": 1,
    "downscale_interval_seconds": 600,
    "autoscale_metrics": [
      {
        "metric": "adapter_queue_time",
        "upscale_threshold": 100,
        "downscale_threshold": 50
      }
    ]
  }
}
```

--------------------------------

### Scale Range Settings Response

Source: https://judoscale.com/docs/api-docs

Example JSON response for scheduled scale range settings. It lists specific times and days with corresponding min/max dyno ranges.

```json
{
  "scale_range_settings": [
    {
      "wday": 2,
      "time": "16:30",
      "range_min": 2,
      "range_max": 4
    },
    {
      "wday": 3,
      "time": "00:00",
      "range_min": 3,
      "range_max": 5
    }
  ]
}
```

--------------------------------

### Example Slack Notification Message

Source: https://judoscale.com/docs/slack-notifications

The raw text format sent by Judoscale when a scaling event occurs.

```text
Judoscale is scaling UP - scaling to 4 web dynos for your-app-production
```

--------------------------------

### GET /process_type/:name/scale_range_settings

Source: https://judoscale.com/docs/api-docs

Fetches the autoscaling schedules for a specific process type.

```APIDOC
## GET /process_type/:name/scale_range_settings

### Description
Fetches the autoscaling schedules (scale range settings) for a specific process type.

### Method
GET

### Endpoint
/process_type/:name/scale_range_settings

### Parameters
#### Path Parameters
- **name** (string) - Required - The name of the process type.

### Response
#### Success Response (200)
- **scale_range_settings** (array) - A list of scale range schedule objects.

#### Response Example
{
  "scale_range_settings": [
    {
      "wday": 2,
      "time": "16:30",
      "range_min": 2,
      "range_max": 4
    }
  ]
}
```

--------------------------------

### Get Process Type Details

Source: https://judoscale.com/docs/api-docs

Retrieve the current autoscaling configuration for a specific process type. This includes settings like enabled status, scale range, and metric thresholds.

```http
GET /process_type/web
```

--------------------------------

### Get Scale Range Settings

Source: https://judoscale.com/docs/api-docs

Retrieve the scheduled scale range settings for a specific process type. This allows for dynamic adjustments to min/max dynos based on time and day.

```http
GET /process_type/web/scale_range_settings
```

--------------------------------

### Enable Busy Job Tracking for RQ

Source: https://judoscale.com/docs/long-running-jobs

Initialize the RQ integration with Judoscale and enable busy job tracking. Ensure you are using the latest package version.

```python
judoscale_rq(redis, extra_config={\"RQ\": {\"TRACK_BUSY_JOBS\": True}})
```

--------------------------------

### Open Judoscale Add-on via Heroku CLI

Source: https://judoscale.com/docs/heroku-getting-started

This command opens the Judoscale add-on in your browser after it has been provisioned. Replace YOUR_APP_NAME with your Heroku app name.

```bash
heroku addons:open judoscale -a YOUR_APP_NAME
```

--------------------------------

### Enable Busy Job Tracking for Sidekiq

Source: https://judoscale.com/docs/long-running-jobs

Configure the Judoscale adapter to report the number of actively running jobs for each queue. Ensure you are using the latest package version.

```ruby
config.sidekiq.track_busy_jobs = true
```

--------------------------------

### Fetch App Settings with curl

Source: https://judoscale.com/docs/api-docs

Retrieves current application settings using the JUDOSCALE_URL environment variable.

```bash
export JUDOSCALE_URL=$(heroku config:get JUDOSCALE_URL)
curl $JUDOSCALE_URL/settings
```

--------------------------------

### Retrieve Application Settings

Source: https://judoscale.com/docs/api-docs

Request the current application settings snapshot.

```http
GET /settings
```

--------------------------------

### Validation Error Response

Source: https://judoscale.com/docs/api-docs

Example of a validation error response (422 status code) when updating scale range settings, indicating issues with the provided values.

```json
{
  "errors": [["exceeds the plan limit"], []]
}
```

--------------------------------

### Scale Range Settings Update Request Body

Source: https://judoscale.com/docs/api-docs

JSON payload for updating scale range settings. This example modifies the range for a specific time and day.

```json
{
  "scale_range_settings": [
    {
      "wday": 2,
      "time": "16:30",
      "range_min": 2,
      "range_max": 99
    },
    {
      "wday": 3,
      "time": "00:00",
      "range_min": 3,
      "range_max": 5
    }
  ]
}
```

--------------------------------

### Enable Busy Job Tracking for Celery

Source: https://judoscale.com/docs/long-running-jobs

Initialize the Celery integration with Judoscale and enable busy job tracking. Ensure you are using the latest package version.

```python
judoscale_celery(celery_app, extra_config={\"CELERY\": {\"TRACK_BUSY_JOBS\": True}})
```

--------------------------------

### Register Custom Adapter with Judoscale Reporter

Source: https://judoscale.com/docs/custom-worker-python

Add a custom adapter to the Judoscale reporter. This involves creating an instance of your custom metrics collector, initializing an `Adapter` with a unique name and adapter info, and then adding it to the reporter. Ensure `judoscale.core.adapter`, `judoscale.core.config`, and `judoscale.core.reporter` are imported.

```Python
from judoscale.core.adapter
from judoscale.core.config
from judoscale.core.reporter

collector = FileSystemQueueMetricsCollector(config=judoconfig)
adapter = Adapter(
    "judoscale-filesystem",
    adapter_info=AdapterInfo(platform_version="0.0.1"),
    metrics_collector=collector,
)
reporter.add_adapter(adapter)

```

--------------------------------

### Process Type Update Request Body

Source: https://judoscale.com/docs/api-docs

JSON payload for updating a process type's autoscaling configuration. This example enables autoscaling and sets basic thresholds and range.

```json
{
  "process_type": {
    "name": "web",
    "autoscaling_enabled": true,
    "range_min": 2,
    "range_max": 10,
    "upscale_threshold": 100,
    "downscale_threshold": 50
  }
}
```

--------------------------------

### Configure the Custom Adapter

Source: https://judoscale.com/docs/custom-worker-ruby

Sets optional configuration parameters for the adapter. This block must be placed after the add_adapter call.

```ruby
Judoscale.config do |config|
  config.file_system_queue.max_queues = 50
end
```

--------------------------------

### Upgrade Heroku Add-on Plan

Source: https://judoscale.com/docs/pricing-details

Use this command to upgrade your Judoscale Heroku add-on to a specific 'belt' plan. Replace 'purple' with your desired plan (e.g., 'blue'). This change affects plan validation for dyno scaling on the Judoscale side.

```bash
heroku addons:upgrade judoscale:purple -a your-app-name
```

--------------------------------

### Provision Judoscale Add-on via Heroku CLI

Source: https://judoscale.com/docs/heroku-getting-started

Use this command to add the Judoscale add-on to your Heroku application. Replace YOUR_APP_NAME with your actual Heroku app name.

```bash
heroku addons:create judoscale -a YOUR_APP_NAME
```

--------------------------------

### Set JUDOSCALE_URL Environment Variable

Source: https://judoscale.com/docs/heroku-addon-migration

Configure the new environment variable required for the direct integration to report metrics.

```bash
heroku config:set JUDOSCALE_URL=https://adapter.judoscale.com/... --app your-app-name
```

--------------------------------

### Dockerfile for NGINX Sidecar

Source: https://judoscale.com/docs/nginx-sidecar

Use this Dockerfile to build a minimal NGINX sidecar container. It copies a custom NGINX configuration into the container.

```dockerfile
FROM nginx:latest

COPY nginx.conf /etc/nginx/conf.d/default.conf

```

--------------------------------

### Register the Custom Adapter

Source: https://judoscale.com/docs/custom-worker-ruby

Registers the new adapter with the Judoscale core gem, specifying the metrics collector class and configuration exposure.

```ruby
Judoscale.add_adapter :"judoscale-file-system-queue", {},
  metrics_collector: Judoscale::FileSystemQueue::MetricsCollector,
  # This is the name that gets exposed to `Judoscale.config`
  expose_config: Judoscale::Config::JobAdapterConfig.new(:file_system_queue)
```

--------------------------------

### Judoscale Node.js Version Requirements

Source: https://judoscale.com/docs/understanding-utilization

Ensure you are using the correct version of judoscale-fastify (2.3.0+) or judoscale-express (2.2.0+) to collect utilization metrics.

```javascript
judoscale-node: `judoscale-fastify` version `2.3.0` or greater, `judoscale-express` version `2.2.0` or greater;
```

--------------------------------

### List Services Endpoint Response

Source: https://judoscale.com/docs/custom-platform-getting-started

Sample JSON response containing service details.

```json
[
  { "id": "web-stag", "group": "custom-app-staging", "scale": 2, "web": true },
  { "id": "worker-stag", "group": "custom-app-staging", "scale": 1 },
  { "id": "web-prod", "group": "custom-app-production", "scale": 5, "web": true },
  { "id": "worker-prod", "group": "custom-app-production", "scale": 2 }
]
```

--------------------------------

### Set JUDOSCALE_URL environment variable

Source: https://judoscale.com/docs/fly-getting-started

Set the JUDOSCALE_URL as a secret in your Fly app. This is required for the Judoscale package to send metrics to the correct service. Do not define this in a Dockerfile or .env file.

```bash
fly secrets set JUDOSCALE_URL="https://example.judoscale.com/v3/adapter/."
```

--------------------------------

### Implement a Custom Metrics Collector

Source: https://judoscale.com/docs/custom-worker-ruby

Defines the metrics collection logic by implementing .adapter_config and #collect. The collector must return an array of Metric objects representing queue latency.

```ruby
# config/initializers/judoscale.rb

module Judoscale
  module FileSystemQueue
    class MetricsCollector < Judoscale::JobMetricsCollector
      def self.adapter_config
        # This comes from the add_adapter call below
        Judoscale::Config.instance.file_system_queue
      end

      def collect
        metrics = []
        t = Time.now.utc

        Dir["tmp/queues/*"].each do |queue_path|
          queue = queue_path.split("/").last

          # Queue time is based on the oldest job in the queue
          job_enqueued_times = Dir["#{queue_path}/*"].map { File.new(_1).birthtime }
          oldest = job_enqueued_times.min
          queue_time_ms = oldest ? ((t - oldest) * 1000).ceil : 0

          # This is the important part!
          metrics.push Metric.new(:qt, queue_time_ms, t, queue)
        end

        # Set log level to DEBUG to see these logs
        log_collection(metrics)

        # Also very important to return the metrics
        metrics
      end
    end
  end
end
```

--------------------------------

### Judoscale Ruby Version Requirement

Source: https://judoscale.com/docs/understanding-utilization

Ensure you are using version 1.12.0 or greater of the judoscale-ruby gem to collect utilization metrics.

```ruby
judoscale-ruby: version `1.12.0` or greater;
```

--------------------------------

### NGINX Configuration for Header Injection

Source: https://judoscale.com/docs/nginx-sidecar

This NGINX configuration listens on port 80, adds the X-Request-Start header with the current millisecond timestamp, and proxies requests to the application container running on localhost:3000.

```nginx
server {
     listen 80;

     location / {
         proxy_set_header X-Request-Start "t=${msec}";

         proxy_pass http://localhost:3000;
     }
 }

```

--------------------------------

### Judoscale Python Version Requirement

Source: https://judoscale.com/docs/understanding-utilization

Ensure you are using version 1.10.0 or greater of the judoscale-python package to collect utilization metrics.

```python
judoscale-python: version `1.10.0` or greater;
```

--------------------------------

### CloudFormation Template for Judoscale IAM Role

Source: https://judoscale.com/docs/aws-iam-notes

This CloudFormation script creates an IAM role and a managed policy for Judoscale, granting read-only access to ECS resources and account metadata. It includes parameters for role name, Judoscale account ID, and an external ID for secure cross-account role assumption.

```json
{
  "AWSTemplateFormatVersion": "2010-09-09",
  "Description": "Create an IAM role for Judoscale. Grants read-only access to all ECS resources and some account metadata.",
  "Parameters": {
    "RoleName": {
      "Type": "String",
      "Description": "IAM Role Name"
    },
    "JudoscaleAccountId": {
      "Type": "String",
      "Description": "Judoscale Account ID"
    },
    "ExternalId": {
      "Type": "String",
      "Description": "The unique ID for your team provided by Judoscale"
    }
  },
  "Resources": {
    "JudoscaleIAMRole": {
      "Type": "AWS::IAM::Role",
      "Properties": {
        "RoleName": { "Ref": "RoleName" },
        "AssumeRolePolicyDocument": {
          "Version": "2012-10-17",
          "Statement": [
            {
              "Effect": "Allow",
              "Principal": {
                "AWS": {
                  "Fn::Sub": "arn:aws:iam::${JudoscaleAccountId}:root"
                }
              },
              "Action": ["sts:AssumeRole"],
              "Condition": {
                "StringEquals": {
                  "sts:ExternalId": { "Ref": "ExternalId" }
                }
              }
            }
          ]
        },
        "ManagedPolicyArns": [
          {
            "Ref": "JudoscaleIAMPolicy"
          }
        ]
      }
    },
    "JudoscaleIAMPolicy": {
      "Type": "AWS::IAM::ManagedPolicy",
      "Properties": {
        "Description": "ECS Read-only permissions policy for Judoscale role",
        "PolicyDocument": {
          "Version": "2012-10-17",
          "Statement": [
            {
              "Effect": "Allow",
              "Action": [
                "ecs:Describe*",
                "ecs:List*",
                "iam:ListAccountAliases",
                "account:ListRegions"
              ],
              "Resource": "*"
            }
          ]
        }
      }
    }
  },
  "Outputs": {
    "RoleARN": {
      "Description": "ARN of the Judoscale IAM role",
      "Value": {
        "Fn::GetAtt": ["JudoscaleIAMRole", "Arn"]
      }
    }
  }
}
```

--------------------------------

### Judoscale API Requirements

Source: https://judoscale.com/docs/custom-platform-getting-started

Details on authentication, content type, and status codes expected by the Judoscale API.

```APIDOC
## Judoscale API Requirements

### Authentication
An auth token is required for every request. Use the `Authorization: Bearer <token>` header.

### Content-Type
Requests and responses must use `JSON` format. Use the `Content-Type: application/json` header.

### Status Codes
- `200` Success
- `401` Unauthorized
- `404` Not Found
```
Tokens:
Raw