Infrastructure as Software in Micro Clouds at the Edge

Area of Science:

• Computer Science
• Distributed Systems
• Cloud Computing

Background:

• Edge computing brings cloud services closer to users, enabling new applications.
• Geo-distributed infrastructure deployment presents challenges in resource utilization and location-specific advantages.
• Real-time applications necessitate multi-tier architectures for local preprocessing and cloud-based heavy workloads.

Purpose of the Study:

• To present a model for dynamically organizing geo-distributed edge nodes into micro clouds.
• To enable elastic resource reorganization based on population needs.
• To abstract infrastructure to the software level, facilitating "infrastructure as software" at the edge.

Main Methods:

• Developing a model that organizes geo-distributed nodes into dynamic micro clouds.
• Adapting cloud organization principles for the edge environment.
• Enabling descriptive specification of desired states for automated system management.

Main Results:

• A model capable of organizing geo-distributed nodes into adaptable micro clouds.
• Elasticity achieved through cloud principles tailored for edge environments.
• Abstraction of infrastructure to software, enabling "infrastructure as software".

Conclusions:

• The proposed model effectively organizes edge resources into dynamic micro clouds.
• The system allows for elastic resource reorganization to meet evolving population demands.
• Integration into existing tools could enable cloud providers to offer micro clouds as a service.