Installation

Add a local runner to the Kubiya platform to deploy actions to & use those from within workflows

Minimum Requirements:

These requirements are estimated based on observed usage statistics and recommended practices.

Cluster Configuration

• Minimum of 2 nodes to ensure high availability.

• Each node should have at least 4 vCPU cores and 8 GB RAM.

Autoscaling

• Cluster autoscaling should be enabled to support scaling up to 5 or more concurrent agents. This ensures that additional resources are automatically provisioned as the load increases.

Per Agent Requirements

Each Kubiya runner agent should be allocated the following resources to operate efficiently:

• CPU Request: 600m (milliCPUs)

• CPU Limit: 600m (milliCPUs)

• Memory Request: 1Gi (Gibibytes)

• Memory Limit: 1Gi (Gibibytes)

Local runners are currently supported only from within Kubernetes

You'll need a Kubernetes cluster in order to setup a local runner

From the web interface

1. Go the the Kubiya Management interface

2. Navigate to the Local Runners page

1. Click on Add Local Runner:

1. Provide a name for your local runner, we suggest to choose a meaningful name so you can easily identify the cluster which will host it:

1. After hitting Continue, you should receive a kubectl command to install the operator deployment:

1. Paste the command in your terminal , make sure that you are in the correct Kubernetes context to make sure the manifest is applied on the correct cluster:

1. After the manifest is successfully applied, you can check the runner health on the Local Runners dashboard:

From the Slack App

1. Go to slack to your Kubi app

2. Execute the /manage command

3. Hit on Create Local Runner

1. Hit on Deploy New Runner

1. Provide a meaningful name (could be the name of the cluster/environment/purpose)

1. Copy the command and execute it within your kubernetes cluster

1. Hit the Connect button and you will have Kubiya being able execute actions in your own environment ready for your order