blob: ae4f74a344c44dae2769660a49dc592523a7ab8c [file] [log] [blame]
In this section, we explain the architecture of Apache Marvin-AI. The platform has some specific features, such as:
1. Language Agnostic capabilities: The platform must be able to support R, Python and Scala algorithms
2. Parallelism capacities in many different levels (Eg. GPU, Multi-core and Multi-node)
3. Deploy distributed machine learning models that are able to receive high concurrent traffic and provide response in near real-time
Platform Concept Map (High Level Integration)
.. image:: ../images/concept-map.png
Quality Attributes
For Data Scientists:
* **Interoperability:** To support different programmer languages
* **Usability:** To accelerate and simplify the model creation process
For Administrators:
* **Manageability:** To simplify the distributed deploy/management process
* **Usability:** To support from tiny to intensive loads
For Marvin Developers:
* **Modifiability:** To improve and release new versions constantly
* **Maintainability:** To allow all type of programmers (from beginners to experts) to contribute
Architectural Tactics
.. image:: ../images/architectural-tactics.png
.. image:: ../images/dasfe.png
More details in :ref:`dasfe`.
Context Diagram
.. image:: ../images/context-diagram.png
Execution Flow
.. image:: ../images/execution-flow.png
Executor and Engine
.. image:: ../images/executor-engine.png
Cluster Admin
.. image:: ../images/cluster-admin.png
Deployment Diagram
.. image:: ../images/deployment-diagram.png