About MULTS.

A lightweight fault-tolerant architecture to dynamic manage unreliable virtual machines (transient servers) in a multi-cloud scenario.

Negotiation Layer

A multi-cloud negotiation layer to manage resources between users and cloud providers enables a new breed of trust services where users no longer know which cloud provider is being used to run their applications in a transparent way.

In principle, this bus needs to be platform independent and autonomously manage local and cloud transient servers, without user intervention, choosing an optimal environment to execute user applications based on a set of user requirements, including local and remote resources in a reliable fault-tolerant environment to guarantee application execution.

Macro Architecture

Our architecture use machine learning to predict revocation events, being useful to define fault tolerance parameters. This is a key step toward successful and efficient usage of these instances to provide trusted services with minimal interruptions at cheapest prices.

Currently, our approach compiles and analyses Amazon EC2 Spot Instances’ historic price change data to support decisions.

Experiments indicates that this model can be used under realistic working conditions with better use of idle resources.