---
title: Custom model Portable Prediction Server
description: How to download, build, and run the custom model Portable Prediction Server (PPS) to deploy a custom model to an external prediction environment.

---

# Custom model Portable Prediction Server {: #custom-model-portable-prediction-server }

The custom model Portable Prediction Server (PPS) is a solution for deploying a custom model to an external prediction environment. It can be built and run disconnected from main installation environments. The PPS is available as a downloadable bundle containing a deployed custom model, a custom environment, and the monitoring agent. Once started, the custom model PPS installation serves predictions via the DataRobot REST API.

## Download and configure the custom model PPS bundle {: #download-and-configure-the-custom-model-pps-bundle }

The custom model PPS bundle is provided for any custom model tagged as having an [external prediction environment](pred-env) in the deployment inventory.

!!! note
	 Before proceeding, note that DataRobot supports Linux-based prediction environments for PPS. It is possible to use other Unix-based prediction environments, but only Linux-based systems are validated and officially supported.

Select the custom model you wish to use, navigate to the **Predictions > Portable Predictions** tab of the deployment, and select **Download portable prediction package**.

![](images/custom-pps-1.png)

Alternatively, instead of downloading the contents in one bundle, you can download the custom model, custom environment, or the monitoring agent as individual components.

![](images/custom-pps-2.png)

After downloading the .zip file, extract it locally with an unzip command:

`unzip <cm_pps_installer_*>.zip`

Next, access the installation script (unzipped from the bundle) to build the custom model PPS with monitoring agent support. To do so, run the command displayed in step 2:

![](images/custom-pps-3.png)

For more build options, such as the ability to skip the monitoring agent Docker image install, run:

`bash ./cm_pps_installer.sh --help`

If the build passes without errors, it adds two new Docker images to the local Docker registry:

* `cm_pps_XYZ` is the image assembling the custom model and custom environment.

* `datarobot/mlops-tracking-agent` is the monitoring agent Docker image, used to report prediction statistics back to DataRobot.

## Make predictions with PPS {: #make-predictions-with-pps }

DataRobot provides two example <a target="_blank" href="https://docs.docker.com/compose/install/">Docker Compose configurations</a> in the bundle to get you started with the custom model PPS:

* `docker-compose-fs.yml`: uses a file system-based spooler between the model container and the monitoring agent container. Recommended for a single model.

* `docker-compose-rabbit.yml`: uses a RabbitMQ-based spooler between the model container and the monitoring agent container. Use this configuration to run several models with a single monitoring agent instance.

!!! note
    To utilize the provided Docker Compose files, be sure you have added the [`datarobot-mlops` package](https://pypi.org/project/datarobot-mlops){ target=_blank } (with additional dependencies as needed) to your model's `requirements.txt` file.

After selecting the configuration to use, edit the Docker Compose file to include the deployment ID and your [API key](api-key-mgmt) in the corresponding fields.

Once configured, start the prediction sever:

* For single models using the file system-based spooler, run:

	`docker-compose -f docker-compose-fs.yml up`

* For multiple models with a single monitoring agent instance, use the RabbitMQ-based spooler:

	`docker-compose -f docker-compose-rabbit.yml up`

When the PPS is running, the Docker image exposes three HTTP endpoints:

* `POST /predictions` scores a given dataset.
* `GET /info` returns information about the loaded model.
* `GET /ping` ensures the tech stack is running.

!!! note
	Note that prediction routes only support comma delimited (CSV) scoring datasets. The maximum payload size is 50 MB.

The following demonstrates a sample prediction request and JSON response:

```
curl -X POST http://localhost:6788/predictions/ \
     -H "Content-Type: text/csv" \
     --data-binary @path/to/scoring.csv
```

	{
		"data": [{
				"prediction": 23.03329917456927,
				"predictionValues": [{
					"label": "MEDV",
					"value": 23.03329917456927
				}],
				"rowId": 0
			},
			{
				"prediction": 33.01475956455371,
				"predictionValues": [{
					"label": "MEDV",
					"value": 33.01475956455371
				}],
				"rowId": 1
			},
		]
	}


###  MLOps environment variables {: #mlops-environment-variables }

The following table lists the MLOps service environment variables supported for all custom models using PPS. You may want to adjust these settings based on the run environment used.

|  Variable  | Description |  Default  |
|------------------------------|--------------|--------------------|
| `MLOPS_SERVICE_URL`  |   The address of the running DataRobot application. | Autogenerated value |
| `MLOPS_API_TOKEN`  |   Your DataRobot API key. | Undefined; must be provided. |
| `MLOPS_SPOOLER_TYPE` |   The type of spooler used by the custom model and monitoring agent.   | Autogenerated value |
| `MLOPS_FILESYSTEM_DIRECTORY` |   The filesystem spooler configuration for the monitoring agent. | Autogenerated value |
| `MLOPS_RABBITMQ_QUEUE_URL`  | The RabbitMQ spooler configuration for the monitoring agent.  | Autogenerated value |
| `MLOPS_RABBITMQ_QUEUE_NAME` |   The RabbitMQ spooler configuration for the monitoring agent.  | Autogenerated value |
| `START_DELAY`  |   Triggers a delay before starting the monitoring agent.                   | Autogenerated value |


###  DRUM-based environment variables {: #drum-based-environment-variables }

The following table lists the environment variables supported for [DRUM-based](custom-local-test) custom environments:

|  Variable | Description | Default |
|----------|---------------|------------------|
| `ADDRESS` | The prediction server's starting address. | `0.0.0.0:6788` |
| `MODEL_ID` | The ID of the deployed model (required for monitoring).   | Autogenerated value |
| `DEPLOYMENT_ID` | The deployment ID.| Undefined; must be provided. |
| `MONITOR` | A flag that enables MLOps monitoring. | True. Provide an empty value or remove this variable to disable monitoring.|
| `MONITOR_SETTINGS` | Settings for the monitoring agent spooler. |  Autogenerated value |

### RabbitMQ service environment variables {: #rabbitmq-service-environment-variables }

| Variable | Description | Default  |
|-------------|-------------------|-----------|
| `RABBITMQ_DEFAULT_USER` | The default RabbitMQ user. | Autogenerated value |
| `RABBITMQ_DEFAULT_PASS` | The default RabbitMQ password. | Autogenerated value |
