Physical Architecture - purpose of "deploy" PCs

Hello all together,
what is the purpose of - or difference between

  • Node PC
    and
  • Deploy Node PC
    and also between
  • Deploy Behavior PC
    and
  • Behavior PC?
    Some “facts” about PCs:
  1. Deploy PCs always appear on the lowest level in the Project Explorer - even if they are created “inside” of another PC - why?
  2. A Behavior PC can only be created inside of a Deploy Behavior PC - not directly inside of a Node PC or Deploy Node PC…
  3. The “lowest” PC must be a Node PC.
  4. A Deploy Behavior PC can be created inside of a Node PC or a Deploy Node PC or inside of another Deploy Behavior PC.
  5. Node PCs can be created inside of Node PCs or Deploy Node PCs
    See picture for some possibilities…

Hello,
As a reminder, some definitions from the Capella Glossary :

  • Behavioral PC : a Physical Component (PC) in charge of implementing / realizing parts of the functions allocated to the system
  • Node PC (aka “implementation” component in Arcadia) : a material Physical Component, resource embedding some behavior components. Node Physical Components deliver resources for behavioral components execution.
    As well as the description of the commands in the Palette :
  • Deploy Behavior PC : Creates a new Physical Component with a Kind BEHAVIOR and deployed on the target Physical Component.
  • Deploy Node PC : Creates a new Physical Component with a Kind NODE and deployed on the target Node Physical Component.
    Deploy PC (both Behavior or Node types) use the resources provided by a PC, in order to perform the functions allocated to them. In the simplest way, this gives the diagram below, with a Deployed Behavioral PC deployed in a Node PC:
    Note that Deploy PC are not contained in PC: deployment relation does not introduce a decomposition of the PC hosting the resources.
    Hope it helps.

Hello Richard,
Semantically, there is only 2 kinds of Physical Components: the Nodes and the Behavior (see definitions provided by Juan above).
In Physical Architecture Blanks [PAB], several tools are provided in order to fasten the operations on the model:

  • Node PC: Creates a new Node PC. If used on a Node PC, uses the relation of decomposition (the new PC is created under the existing one)
  • Deploy Node PC: Creates a new Node PC.
    Need to be used on a Node PC. Uses the relation of deployment(the new PC is created aside the existing one and a deployment link is defined)
  • Deploy Behavior PC: Creates a new Behavior PC.
    Need If used on a Node PC or on Behavior PC. Use the relation of deployment (the new PC is created aside the existing one and a deployment link is defined)
  • Behavior PC: Creates a new Behavior PC. If used on a Behavior PC, uses the relation of decomposition (the new PC is created under the existing one)

Thank you Juan,
I think this was the information I was missing:
“deployment relation does not introduce a decomposition of the PC hosting the resources.”
So a “Node PC” inside another “Node PC” is a decomposition (on the Node-Physical Level).
A “Deploy Behavior PC” is mandatory to place BEHAVIOR/Functions.
A “Behavior PC” inside of a “Deploy Behavior PC” is decomposition (on the Behavior-Physical Level).
What is a “Deploy Behavior PC” inside an other “Deploy Behavior PC” (PC17 inside PC16)?

I would discourage the “Deploy Behavior PC” inside an other “Deploy Behavior PC”
More seriously, I haven’t see this use case, but maybe other users have seen it ?

Deploying a Behavior PC in another Behavior PC should be prohibited by the tool.
It is not compliant with the Arcadia method.

Thanks to you all,
slowly things become more clear - or not…?
So one more Question - What is a “Deploy Node PC”?
“Deploy Node PC: Creates a new Node PC. Need to be used on a Node PC. Uses the relation of deployment(the new PC is created aside the existing one and a deployment link is defined)”
So a Deploy Node PC must be created inside of Node PC - but it is no decomposition of this Node PC, only connected by a semantic relationship-link “Deployed Part” to the Node PC where it was created.
Every Deploy Node PC is placed on the lowest level (in the project explorer) aside all other Deploy Node PCs and the lowest level Node PCs.
So physically - what is a “Deploy Node PC”? Can anyone give an Example?

Deploying a Node PC in another Node PC is a kind of legacy feature of Capella. It is not compliant with Arcadia.
If you really want an example of how it was intended to be used: take the example of a server (Node PC) located in a room or geographical zone (which could also be represented by another Node PC). The idea was that this kind of relationship was weaker than a true containment relationship. But I recommend you forget that and stick to Arcadia

Copyright © Eclipse Capella, the Eclipse Capella logo, Eclipse and the Eclipse logo are Trademarks of The Eclipse Foundation.