The Media Cloud and Microservice Architecture (MCMA) is a project of the EBU MIM-AI (Media Information Management and Artificial Intelligence) strategic programme.

The MCMA (Media Cloud and Microservice Architecture) builds on previous work in the FIMS (Framework for Interoperable Media Services) project.

The MCMA objective is to develop a set of simplified REST APIs with minimum payload to allow integrating workflows combining microservices in the cloud with other in-house services and processes.

MCMA will also share libraries containing glue code between the high level APIs and low level (e.g. cloud) platforms. The MCMA motto is "implementation by the example". MCMA will also publish guidelines to implement (cloud) mediaservice arhcitectures.

The MCMA cloud services provides an abstraction on top of cloud provider specific services allowing cross cloud provider compatibility. Instead of enforcing a strongly type set of interfaces like FIMS 1.X, the new approach leverages a much more open integration model by supporting late bidding object definition (JsonLD) and micro-services architecture. The MCMA libraries also provide a turn on key solution for cloud infrastructure implementation, deployment and management. Concept of service registration and automatic discovery also key features of the infrastructure.  All libraries and components are available as projects hosted on GitHub FIMS repo.   

The MCMA REST calls represent the highest level of abstraction across any of the existing cloud platforms (or else). The libraries, specific to each platform, form the API layer between the REST interface and the tools. FIMS uses the respective cloud infrastructures to facilitate service discovery. Payload such as data returned by AI tools is itself accessed through dedicated REST calls or directly by applications. Guidelines are there to explain how to deploy MCMA. This has been successfully demonstrated at IBC and NAB in 2017.

MCMA is following the developments around production in the cloud where interfaces like ingest, transfer, transform, repository management and QA will play a key role.

On this basis, FIMS proposes to communicate on most recent developments as follows:

  • Continue to enrich github with new libraires and guidelines on
  • Use the MCMA EBU workspace and information page to publish recent results and highlight the technical approach and it benefits with links to github
  • Communicate on MCMA applied to AI and cloud micro-services to convey the message around service orientated architectures and their growing importance in the future (tech-I and other EBU T&I communication tools).

How can EBU T&I help?

EBU is here to help experts meet, share and learn about the cloud, artificial intelligence and microservice architectures.

Guidance & tools