node_role.md

Node Role APIs

Node Roles are the core of OpenCrowbar deployment and orchestration engine

There are four types of data that OpenCrowbar tracks, three of them are maintained on related NodeRoleDatum mode.

1. user data (node_role.data) is set by users during the proposed state (also known as "out bound data")
2. system data (node_role.sysdata) is set by crowbar before annealing (also known as "out bound data")
3. wall data (node_role.wall) is set by the jig after transistion (also known as "in bound data")
4. discovery data (node.wall) is stored on the node instead of node role because it reflects node information aggregated from all the jigs. This information is available using the node.attrib_[name] and Attrib model. Please see the node API docs for more about this type of data

NodeRole does not have a natural key, so you must reference them them by ID or under the relevant Nodes, Roles, or Deployment.

API Actions

Verb	URL	Comments
GET	api/v2/node_roles	List
GET	api/v2/node_roles/:id	Specific Item
PUT	api/v2/node_roles/:id	Update Item
POST	api/v2/node_roles	Create Item
GET	/api/v2/node_roles/[:node_role_id]/attribs	List Attribs for a specific node_role
GET	/api/v2/node_roles/[:node_role_id]/attribs/[:id]	Show Attrib (including value) for a specific Node_Role
PUT	/api/v2/node_roles/[:node_role_id]/attribs/[:id]	Update Attrib
DELETE	-	NOT SUPPORTED

Helpers have been added to NodeRole create so that you do not need to provide IDs when creating a new NodeRole. You can pass:

• Deployment Name instead of Deployment ID
• Node Name instead of Node ID
• Role Name instead of Role ID

JSON fields

Attribute	Type	Settable	Note
Available	Boolean	Yes
Cohort	Integer	??
Created_at	String	No	Unicode - date format
Updated_at	String	No	Unicode - date format
Runlog	String	??
Order	Integer	??
State	Integer	??
Node_Id	Integer	Yes
Status	??	??
Run_count	Integer	No
Deployment_Id	Integer	??
Role_Id	Integer	Yes
Id	Internal Ref	??	Actually an Int