Kore itself runs inside a Kubernetes cluster, and this section discusses how you can use a local instance of Kore on your workstation to bootstrap a cloud cluster to install Kore for your organization.
To run Kore locally, you will need Docker installed. Both Mac OSX and Linux have been tested; this process may not work correctly on a Windows workstation.
For instructions, see Get the CLI.
To get Kore running locally, simply run:
This will take several minutes to create a cluster using Kind and install the current Kore Helm chart into it.
As you have not configured an IDP, you should log in using the credentials output by the above command.
You can also log in to the UI using the same credentials at http://localhost:3000.
Kore needs a cloud account to use in order to create a cluster to run Kore in - find out about the various types of cloud accounts supported by Kore in the cloud accounts section.
Add a cloud account to your local Kore for the cloud you wish to host your Kore cluster in. You can do this via the UI (Configure > Cloud > Organization or Configure > Cloud > Account/Project/Subscription) or using the CLI:
Now you have credentials in your local Kore instance, you can proceed to create the cluster in your chosen cloud provider:
- Create a team (e.g. myorg-infra) in Kore via CLI or via the UI.
- Create the cluster - we recommend doing this via the UI for this initial cluster so you have full visibility of the options used, however it is also possible via the CLI.
Finally, once you have created the cluster successfully, from the CLI, you can add it to your local kubectl contexts, ready to install Kore:
Ensure you choose to set your current context to the cluster you have built.
You now have a Kubernetes cluster to install Kore into - ensure you follow the remaining steps from the prerequisites before continuing to install Kore into this new cluster.