arch_generic_ds

The "generic" Datastore Mode

The "simple" datastore configration mode is limited to only 3 basic datastores (IDs 0, 1, 2):

ds:
  mode: ssh
  mounts:
  - type: system
    path: /mnt/one_datastores/system/
  - type: image
    path: /mnt/one_datastores/default/
  - type: files
    path: /mnt/one_datastores/files/

Fortunately, it's possible to define datastores more freely with the "generic" mode, for example:

ds:
  mode: generic
  config:
    SYSTEM_DS:
      system:
        enabled: false
      system1:
        id: 100
        symlink:
          groups: [node]
          src: /mnt/nfs1/100/
        template: &template
          TYPE: SYSTEM_DS
          TM_MAD: shared
          BRIDGE_LIST: "{{ groups.node | map('extract', hostvars, ['ansible_host']) | join(' ') }}"
      system2:
        id: 101
        symlink:
          groups: [node]
          src: /mnt/nfs2/101/
        template: *template
      system3:
        id: 102
        symlink:
          groups: [node]
          src: /mnt/nfs3/102/
        template: *template
    IMAGE_DS:
      default:
        symlink: { src: /mnt/nfs0/1/ }
        template:
          TM_MAD: shared
    FILE_DS:
      files:
        symlink: { src: /mnt/nfs0/2/ }

With the "generic" mode it's actually possible to configure any number of user-defined datastores, for example:

    SYSTEM_DS:
      system1:
        id: 100
        symlink:
          groups: [node]
          src: /mnt/nfs1/100/
        template: &template
          TYPE: SYSTEM_DS
          TM_MAD: shared
          BRIDGE_LIST: "{{ groups.node | map('extract', hostvars, ['ansible_host']) | join(' ') }}"

Where:

• system1 is a name assigned to the new user-defined datastore
• id is a predicted or existing datatore ID (OpenNebula always starts numbering user-defined datastores from 100)
• symlink is optional, if defined then groups should be used to specify where the actual symlinking has to happen, src specfies the source directory that is going to be symlinked as /var/lib/one/datastores/<id> (unless it's /var/lib/one/datastores/, then it must be skipped)
• template is an usual raw (expressed in YAML) OpenNebula template for the datastore

Warning

When creating user-defined datastores you must provide predicted datastore IDs starting from 100, then incrementing by 1 for subsequent ones. Or, if datatores already exist you can tell one-deploy to start/continue managing them by providing existing datastore IDs.

For the reference, the default "generic" mode configuration (which can be considered an equivalent for the "simple" mode) is:

SYSTEM_DS:
  system:
    id: 0
    managed: true
    enabled: true
    symlink:
      groups: [node]
      src: /var/lib/one/datastores/ # this skips symlinking
    template:
      TYPE: SYSTEM_DS
      TM_MAD: ssh
IMAGE_DS:
  default:
    id: 1
    managed: true
    symlink:
      groups: [frontend, node]
      src: /var/lib/one/datastores/ # this skips symlinking
    template:
      TYPE: IMAGE_DS
      DS_MAD: fs
      TM_MAD: ssh
FILE_DS:
  files:
    id: 2
    managed: true
    symlink:
      groups: [frontend]
      src: /var/lib/one/datastores/ # this skips symlinking
    template:
      TYPE: FILE_DS
      DS_MAD: fs
      TM_MAD: ssh