Pre-deployment preparations

The following preparatory steps should be completed before starting the deployment procedure for any type of Plixer ML Engine appliance.

Deploying the Plixer Machine Learning VM

Use the template obtained with the Plixer One Enterprise license to deploy the Plixer ML VM locally.

This VM will function as a separate deployment host and includes all prerequisite resources. The Plixer ML Engine environment will be deployed and managed from this VM.

Note

For vSphere multi-node deployments, the template should be added to vSphere. See the vSphere multi-node cluster deployment guide for more information.

License and engine registration

Creating an authentication token

The Plixer ML Engine will require an authentication token for Plixer Scrutinizer, which can be created as follows:

  1. In the Plixer Scrutinizer web interface, navigate to Admin > Resources > ML Engines.

  2. Click the + button to add a new ML engine.

  3. From the dropdown, select the type of ML engine paired with your Plixer Scrutinizer environment:

  4. Enter a name to assign to the engine, and then click Save.

After returning to the main view, click the name of the new ML engine and save/copy the primary reporter address and authentication token shown in the tray. These will be required during the Plixer ML Engine deployment process.

Confirming SSH credentials

To complete the appliance setup process, the Plixer ML Engine will need to establish an SSH session with the primary reporter/server of the Plixer Scrutinizer environment. As such, the IP address of the primary reporter and the plixer user password will need to be provided.

If a private SSH key is required, verify that the public key is configured on the primary Plixer Scrutinizer reporter/server under /home/plixer/.ssh/authorized keys. The private key should also be accessible from the machine hosting the Plixer ML Engine virtual appliance, as the path to the key will need to be entered during the appliance setup process.