Blueprint Resources

Resources are the infrastructure building blocks you provision in a blueprint. Teabar supports a wide variety of resource types, from Kubernetes clusters and VMs to Helm releases and DNS entries.

Resource Types

Resources Section Overview

All resources are defined under spec.resources in your blueprint:

apiVersion: teabar.dev/v1
kind: Blueprint
metadata:
  name: my-environment

spec:
  resources:
    # Kubernetes clusters
    clusters:
      - name: main
        provider: hetzner
        type: talos
        # ...

    # Virtual machines
    vms:
      - name: dev-server
        provider: hetzner
        # ...

    # Helm releases
    helm:
      - name: gitlab
        cluster: main
        chart: gitlab/gitlab
        # ...

    # Kubernetes manifests
    manifests:
      - name: config
        cluster: main
        template: |
          # YAML manifest here

    # DNS entries
    dns:
      - name: app
        type: A
        # ...

    # Secrets
    secrets:
      - name: api-key
        type: generated
        # ...

Resource Dependencies

Resources can reference each other using template expressions:

spec:
  resources:
    clusters:
      - name: main
        provider: hetzner
        type: talos
        nodes:
          workers: 3

    helm:
      - name: ingress
        cluster: main  # Reference cluster by name
        chart: ingress-nginx/ingress-nginx

    dns:
      - name: app
        type: A
        # Reference the cluster's ingress IP
        target: "{{ .Resources.clusters.main.ingress_ip }}"

Teabar automatically resolves dependencies and provisions resources in the correct order.

Common Patterns

Per-Participant Resources

Use the count field with the participant index to create resources for each participant:

spec:
  variables:
    - name: participant_count
      type: integer
      default: 10

  resources:
    vms:
      - name: "workstation-{{ .Index }}"
        count: "{{ .Variables.participant_count }}"
        provider: hetzner
        type: cx21
        image: ubuntu-22.04

This creates workstation-0, workstation-1, …, workstation-9 VMs.

Conditional Resources

Enable or disable resources based on variables:

spec:
  variables:
    - name: enable_monitoring
      type: boolean
      default: false

  resources:
    helm:
      - name: prometheus
        cluster: main
        chart: prometheus-community/kube-prometheus-stack
        enabled: "{{ .Variables.enable_monitoring }}"

Shared vs. Isolated Resources

Some resources are shared across participants, while others are created per-participant:

spec:
  resources:
    # Shared: One GitLab instance for all participants
    helm:
      - name: gitlab
        cluster: main
        namespace: gitlab
        chart: gitlab/gitlab

    # Isolated: One namespace per participant
    manifests:
      - name: participant-ns
        cluster: main
        count: "{{ .Variables.participant_count }}"
        template: |
          apiVersion: v1
          kind: Namespace
          metadata:
            name: participant-{{ .Index }}

Resource Lifecycle

Resources follow the environment lifecycle:

1. Creation - Resources are provisioned when the environment starts
2. Updates - Some resources can be updated in-place; others require recreation
3. Checkpoints - Resource state can be captured for quick restoration
4. Destruction - Resources are cleaned up when the environment is destroyed or expires

Provider Support

Different resource types support different providers:

Next Steps