Section: New Results

Distributed Cloud Computing

Participants : Teodor Crivat, Yvon Jégou, Vlad Mirel, Christine Morin, Anne-Cécile Orgerie, Edouard Outin, Nikolaos Parlavantzas, Jean-Louis Pazat, Guillaume Pierre, Aboozar Rajabi, Carlos Ruiz Diaz, Arnab Sinha, Genc Tato, Cédric Tedeschi.

A multi-objective adaptation system for the management of a Distributed Cloud

Participants : Yvon Jégou, Edouard Outin, Jean-Louis Pazat.

In this project, we consider a “Distributed Cloud” made of multiple data/computing centers interconnected by a high speed network and belonging to the same administration domain. Moreover, in the Cloud organization targeted here, the network capabilities can be dynamically configured in order to guarantee QoS for streaming or to negotiate bandwidth for example.

As a first step, we are focusing on a single centralised Cloud.

Due to the dynamic capabilities of the Clouds, often referred to as elasticity, there is a strong need to dynamically adapt both platforms and applications to users needs and environmental constraints such as electrical power consumption.

We address the management of a Cloud in order to consider both optimization for energy consumption and for users’ QoS needs. The objectives of this optimization will be negotiated as contracts on Service Level Agreement (SLA). A special emphasis will be put on the distributed aspect of the platform and include both servers and network adaptation capabilities.

The design of the system relies on self-* techniques and on adaptation mechanisms at any level (from IaaS to SaaS). The MAPE-k framework (Monitor-Analysis-Planning-Execution based on knowledge) is used for the implementation of the system. The technical developments are based on the Openstack framework.

We have implemented a system that uses a genetic algorithm to optimize Cloud energy consumption and machine learning techniques to improve the fitness function regarding a real distributed cluster of servers. We have carried out experiments on the OpenStack platform to validate our solution. This experimentation shows that the machine learning produces an accurate energy model, predicting precise values for the simulation.

We are currently refining this model and comparing it to real measurements on the platform.

This work is done in cooperation with the DIVERSE team and in cooperation with Orange under the umbrella of the B-COM Technology Research Center.

Dynamic reconfiguration for multi-cloud applications

Participants : Nikolaos Parlavantzas, Aboozar Rajabi, Carlos Ruiz Diaz, Arnab Sinha.

In the context of the PaaSage European project, we are working on model-based, continuous self-optimization of multi-cloud applications. In particular, we are developing a dynamic adaptation system, capable of transforming the currently running application configuration into a target configuration in a cost-effective and safe manner. In 2015, we have improved and extended the Adapter prototype  [45] . The system now fully supports dynamic configuration, including detecting changes, generating reconfiguration plans, validating plans based on a cost-benefit calculation, and executing plans in parallel, improving adaptation performance. Moreover, we have performed initial investigations on the use of PaaSage for supporting Internet of Things (IoT) applications [27] . Finally, in the context of Carlos Ruiz's stay, we are defining a model for managing the configuration of cloud applications and environments. This model is based on feature modeling and the derived configurations are mapped to PaaSage models.

Towards a distributed cloud inside the backbone

Participants : Christine Morin, Anne-Cécile Orgerie, Genc Tato, Cédric Tedeschi.

The DISCOVERY proposal officially started at the end of 2015. It is an Inria Project Lab (IPL) led by Adrien Lebre from the ASCOLA team, and currently on leave at Inria. It aims at designing a distributed cloud, leveraging the resources we can find in the network backbone. (The DISCOVERY website: http://beyondtheclouds.github.io ) In practice, this work is intended to get integrated within the OpenStack softwarehttps://www.openstack.org/ so as to decentralize its whole architecture.

In this context, and in collaboration with ASCOLA and ASAP teams, we started the design of an overlay network whose purpose is to be able, with a limited cost, to locate geographically-close nodes from any point of the network. In this framework, the PhD thesis of Genc Tato started in December 2015. It aims at developing locality mechanisms at the data management layer.

We have also started an energy/cost-benefit analysis of a decentralized Cloud infrastructure like the one proposed within Discovery. This work is conducted by Anthony Simonet, a post-doctoral researcher on an Inria contract for the Discovery IPL and co-supervised by Adrien Lebre from the ASCOLA team and Anne-Cécile Orgerie from Myriads team.

Mobile edge cloud computing with ConPaaS

Participants : Teodor Crivat, Vlad Mirel, Guillaume Pierre.

Interactive multi-user applications usually rely on intermediate cloud servers to mediate the inter-user interaction. However, current mobile networks exhibit network latencies in the order of 50-150 ms between the device and any cloud. Such latencies make it impossible to create smooth interactions with the end user. To enable an “instantaneous” feeling, augmented reality applications require that end-to-end latencies should remain below 20 ms.

To address these issues, we extended ConPaaS to support the deployment of cloud applications in a distributed set of Raspberry Pi machines. The motivation is to reduce the latency compared to a traditional deployment where the backend is located in an external cloud: instead of reaching the cloud through a wide-area network, in this setup each cloud node is also equipped with a wifi hotspot which allows local users to access it directly.

Fog Computing

Participant : Jean-Louis Pazat.

The concept of “Fog Computing” is currently developed on the idea of hosting instances of services not on centralized datacenters (i.e. the “Cloud”), but on a highly distributed infrastructure: the Internet Edge (i.e. the “Fog”). This infrastructure consists in geographically distributed computing resources with relatively small capabilities. Compared with datacenters, a “Fog” infrastructure is able to offer to Service Providers a shorter distance from the service to the user but with the same flexibility of software deployment and management.

This work focus on the problem of resource allocation in such infrastructure when considering services in the area of Internet of Things, Social Networks or Online Gaming. For such use-cases, service-to-user latency is a critical parameter for the quality of experience. Optimizing such a parameter is an objective for the platform built on top of the Fog Infrastructure that will be dedicated to the deployment of the considered service. In order to achieve such a goal, the platform needs to select some strategies for the allocation of network and computing resources, based on the initial requirements for service distribution.

We are designing a prototype based on micro services and we are considering low overhead virtualization systems using containers. This prototype is intended to run inside an Internet Box or inside a LAN disk server at user's home. The whole system will be intended to be used very small or medium size user communities willing to share devices and data. The main characteristics of the system will be reliable distributed storage and distributed execution of services.

This work is part of Bruno Stevant's PhD thesis, which began in December 2014. It is done in cooperation with the REOP team, Institut Mines telecom/IRISA.