Page History

...

Placing this information in the routing key provides the ability to enforce access control at class, operation, and method granularity. It also separates the command structure from the content of the management message (i.e. element values) allowing the content to be encrypted and signed end-to-end while still allowing access control at the message-transport level. This means that special access control code need not be written for the management agent.
There are two general types of routing/binding key:

• Command messages simply use the key: agent
• Method keys have the structure: agent.method.<package>.<class>.<method> where
  • <package> is the namespace in which the <class> name is valid,
  • <class> is the name of the class as defined in the schema, and
  • <method> is one of "get", "set", or a schema-defined class-specific method name.
• Unsolicited keys have the structure: mgmt.<agent>.<type>.<package>.<class>.<severity> where
  • <agent> is the uuid of the originating management agent,
  • <type> is one of "schema", "prop", "stat", or "event",
  • <package> is the namespace in which the <class> name is valid, and
  • <class> is the name of the class as defined in the schema.
  • <severity> is relevant for events only. It is one of "critical", "error", "warning", or "info".

...