Advanced Cluster Setup

Deploy Multiple Clusters#

Having one root orchestrator and one cluster orchestrator on one device is a great way to start using Oakestra, but the true power of the system lies in it’s federated architecture.

This guide will walk you through deploying stand-alone components, allowing you to deploy multiple cluster orchestrators across multiple devices. These clusters will be managed by a single root orchestrator.

Stand-alone Root Orchestrator#

First, let’s deploy a stand-alone root orchestrator. This component will manage your clusters and provide an interface for users to interact with the Oakestra setup by providing an API and a dashboard.

curl -sfL oakestra.io/install-root.sh | sh - 

This script will download the required files to the directory ~/oakestra/root_orchestrator. From there it will build the root orchestrator.

Stand-alone Cluster Orchestrator#

Next, we can deploy a stand-alone cluster orchestrator. This component will manage the nodes by delegating applications, creating subnetworks and facilitating communication. Additionally the cluster orchestrator sends aggregated reports to the root orchestrator.

curl -sfL oakestra.io/install-cluster.sh | sh - 

This script will download the required files to the directory ~/oakestra/cluster_orchestrator. From there it will walk you through configuring the cluster. Once the setup is complete, it will build the cluster orchestrator and register with the root orchestrator.

You can register as many cluster orchestrators with the root orchestrator as you would like. Repeat the above command on a new device and configure the orchestrator with a unique Cluster Name and Cluster Location.

Custom Deployments#

Oakestra also allows you to customize your system deployment to best suit your needs.

Configure the Orchestrators#

The root and cluster orchestrator can be configured using environment variables. If they have not been set, they are automatically set when using the above scripts. They can also be manually configured, e.g. when building using the docker compose files.

Root Orchestrator Environment Variables

• SYSTEM_MANAGER_URL: Specify how the root orchestrator can be reached (URL or IP)

# Example configuration for root orchestrator
export SYSTEM_MANAGER_URL=192.158.18.104

Cluster Orchestrator Environment Variables

• SYSTEM_MANAGER_URL: Specify how the root orchestrator can be reached (URL or IP)
• CLUSTER_NAME: Specify the name of the cluster. Make sure this is unique for every additional cluster
• CLUSTER_LOCATION: (optional) Specify the location of the cluster in the format <latitude>, <longitude>, <radius>

# Example configuration for cluster orchestrator
export SYSTEM_MANAGER_URL=192.158.18.104
export CLUSTER_NAME=My_Cluster
export CLUSTER_LOCATION=48.26275365157924,11.668627302307236,100

Choose a Different Branch#

By default these scripts will compose the services of the main oakestra branch, the latest stable release. However this can be changed by setting the environment variable OAKESTRA_BRANCH before running the startup script. This allows you to experiment with some unreleased features.

export OAKESTRA_BRANCH=develop

This will allow you to use the services from the alpha Oakestra release.

Compose Overrides#

Since Oakestra uses docker-compose to build the components, we can use overrides to fine-tune our build environment.

To use the override files, specify the them in a comma-seperated list by setting the OVERRIDE_FILES env variable before running the startup script.

export OVERRIDE_FILES=override-alpha-versions.yaml