Anomaly Detection > Implementing Anomaly Detection > Install Anomaly Detection External Services
  
Install Anomaly Detection External Services
Two external services are required in order to use Anomaly Detection: a Results service and a Training service. Each service can be installed either via a JAR file process or via a Docker process.
Before Installation
Before you install the microservices, either via the JAR file or the Docker process, create a properties file called microservices.properties. You can create the file using any text editor and place the file anywhere that you can reference later in microservice run commands. The file must contain the following properties:
results.destination.uri=results://<results-IPaddress>:<port#>/results/models
This property points to the location of the Results microservice so that the Training microservice can find and store models.
file.storage.path=file:/<path to model storage>
This property defines where the PMML models will be stored when they are generated.
Docker: For a Docker installation, this path can be anywhere. If you use /tmp, model storage will be mapped to the same location as the internal H2 database.
JAR: For a JAR file installation, this path should point to a folder nested in the same location where the .jar and .yml files are stored.
The following three properties do not need to contain any values, however they need to be present in the properties file in order for the microservices to start:
dataset.webservices.url=
dataset.app.id=
dataset.app.key=
The microservices.properties file configuration is a one-time process and the file will be used by both microservices.
JAR Installations
To install via the JAR file process, use the steps outlined in both of the following sections:
Results — JAR Install Process
Training — JAR Install Process
The following video demonstrates the JAR installation process:
Deploying Training and Result Microservices via JAR Files
Docker Installations
To install via the Docker image process, use the steps outline in both of the following sections.
Results — Docker Install Process
Training — Docker Install Process
Before installing either Docker image, ensure that the following prerequisites are met:
You are using an operating system on your host server that supports Docker Engine. Docker is supported on Linux, Cloud, Windows, and OS X systems. See the Install Docker Engine site for specific operating system versions and instructions.
On the Install Docker Engine site, find the installation instructions for your host operating system and follow them to install Docker version 17.03.0-ce or higher.
If you have an older Windows or Mac operating system that cannot run Docker natively, and you have downloaded the Docker Toolbox (see the item above), take the following additional steps.
Increase the amount of RAM assigned to your Docker VM to at least 4 GB (4096 MB).
To increase the RAM, open the settings for your Docker VM from your virtual machine manager application (such as VirtualBox or VM Ware). Increase the system Base Memory to 4096 MB (4 GB).
* 
If your Docker container is already running, stop it before you changing the RAM settings. Open a command terminal and enter docker-machine stop. When the setting changes are complete, return to the command terminal and enter docker-machine start to restart the Docker container.
Share the folder where the microservices.properties file is stored on your host server with the Docker Toolbox.
To share the folder, stop the Docker container, open a virtual machine manager application, add a shared folder for the location of the microservices.properties file, set the new shared folder to auto-mount, and restart the Docker container. For detailed instructions to complete this procedure, see Article - CS266265 on the PTC Community website or view the following video which demonstrates the Docker installation.
Deploying Training and Result Microservices via Docker Containers