Provisioning of node-specific Hieradata¶

Collecting the node UUID¶

The node-specific hieradata is provisioned based on the node UUID, which is hardware dependent and immutable across reboots/reinstalls.

First make sure the introspection data is available for the target node, if it isn’t one can run introspection for a particular node as described in: Introspecting a Single Node.

Then extract the machine unique UUID for the target node with a command like:

openstack baremetal introspection data save NODE-ID | jq .extra.system.product.uuid

where NODE-ID is the target node Ironic UUID. The value returned by the above command will be a unique and immutable machine UUID which isn’t related to the Ironic node UUID. For the next step, we’ll assume the output was 32E87B4C-C4A7-418E-865B-191684A6883B.

Creating the Heat environment file¶

Assuming we want to use /dev/sdc as data disk for ceph-osd on our target node, we’ll create a yaml file looking like the following (eg. my-node-settings.yaml):

resource_registry:
  OS::TripleO::CephStorageExtraConfigPre: /path/to/tripleo-heat-templates/puppet/extraconfig/pre_deploy/per_node.yaml

parameter_defaults:
  NodeDataLookup: |
    {"32E87B4C-C4A7-418E-865B-191684A6883B": {"ceph::profile::params::osds": {"/dev/sdc": {}}}}

In the above example we’re customizing only a single key for a single node, but the structure is that of a UUID-mapped hash so it is possible to customize multiple and different keys for multiple nodes.

Finally, add such an environment file to the deploy commandline:

openstack overcloud deploy [other overcloud deploy options] -e ~/my-node-settings.yaml