Skip to content
OtterScale

Scope

Scopes allow you to organize your resources and control access.

Selecting a Scope

Section titled “Selecting a Scope”

You can select a scope from the dropdown menu in the navigation bar. This will filter the resources you see to only those within the selected scope.

Creating a Scope

Section titled “Creating a Scope”

To create a new scope, navigate to the Scope management area and click on “Create Scope”. A side panel will appear with the creation form.

  1. Select a Plan

    At the top of the form, you will see the details of the plan associated with the scope you are creating. Different plans support different features:

    FeatureStandardAdvancedEnterprise
    Ceph
    Kubernetes
    Cluster
    NodeSingle-NodeMulti-NodeMulti-Node

  2. Configure Scope Details

    Fill in the following information to configure your new scope:

    • Scope Name: Enter a unique name for your scope.
    • Machine: Select the physical machine to assign to this scope.
    • Storage Devices: Select the block devices to be used. (Appears after selecting a machine).
    • Network Configuration (Optional):
      • Calico CIDR: The CIDR block for Calico networking (e.g., 192.168.0.0/16).
      • Virtual IP: The virtual IP address for the scope (e.g., 192.168.1.1).

  3. Create Scope

    Once you have filled in the details, click the Create button to initiate the creation process.

Automated Provisioning

Section titled “Automated Provisioning”

After you initiate the creation, the system automatically performs the following steps:

  1. Create Scope: Initializes the scope record.

    This step maps to the Juju CreateModel operation. It establishes a new isolated workspace (Model) within the Juju controller. Key actions include:

    • Namespace Isolation: Creates a logical boundary for resources, ensuring that applications and machines in this scope do not interfere with others.
    • Environment Configuration: Sets up model-level configurations, such as APT mirrors for faster package downloads.
    • Access Control: Automatically injects SSH keys to grant authorized users access to the machines within this scope.

  2. Add Machine Tags: Tags the selected machine.

    This step applies specific tags to the machine in MAAS. Tags are used to:

    • Mark Ownership: Associates the machine with the specific scope being created.
    • Reserve Resources: Prevents other processes from accidentally using this machine while it is being set up.
    • Define Capabilities: Can be used to indicate specific hardware capabilities or roles (e.g., compute, storage) required for the node.

  3. Commission Machine: Configures the machine (SSH, networking, storage).

    Commissioning is a critical step to verify hardware and prepare it for deployment. It involves:

    • Hardware Inventory: Boots an ephemeral OS to scan and record detailed specs (CPU, Memory, Disks, NICs).
    • Health Checks: Runs scripts to test hardware functionality (Disk I/O, Memory, IPMI/BMC).
    • Configuration: Sets up out-of-band management (BMC/IPMI) and registers network/storage interfaces.
    • Status Change: On success, the machine becomes Ready for deployment. On failure, it is marked as Failed commissioning for troubleshooting.

  4. Wait for Ready: Waits for the machine to report a Ready status.

    The Ready status is mandatory before any further actions can be taken. It confirms that:

    • Commissioning has completed successfully.
    • The machine is free and available in the pool.
    • It is eligible to be allocated and deployed with an operating system. Attempting to deploy a machine that is not Ready (e.g., still commissioning or in a broken state) will fail.

  5. Create Machine: Associates the machine with the scope.

    This step executes the Juju AddMachines operation. It registers the specific physical machine (from MAAS) into the Juju controller. This action triggers:

    • Provisioning: Juju instructs MAAS to deploy the operating system (e.g., Ubuntu) onto the machine.
    • Agent Installation: Once the OS is running, the Juju Agent is installed, allowing Juju to manage the machine as a node.

  6. Wait for Agent: Waits for the agent to start on the machine.

    This step monitors the Juju Agent status until it reports as started. This signifies that:

    • The OS installation is complete and the machine has booted.
    • The Juju Agent is running and has successfully connected to the Juju Controller.
    • The node is now online and ready to accept software deployments (Charms).

  7. Create Node: Finalizes the node creation with network and storage settings.

    This is the final orchestration step where the actual software stack is deployed. It performs the following actions:

    • Network Reservation: Reserves necessary IP addresses for internal services (e.g., Ceph NFS, Kubernetes Load Balancer).
    • Software Deployment: Deploys and configures the core components:
      • Ceph: For distributed storage (using the selected OSD devices).
      • Kubernetes: For container orchestration (configuring Calico networking and Load Balancers).
      • Addons: Additional system utilities.
      • Observability (COS): Monitoring and logging stack.
    • Relation Establishment: Connects these components together to ensure they work as a cohesive system.