Section: New Results

Distributed programming and the Cloud

Participants : Frederico Alvares, Bastien Confais, Simon Dupont, Md Sabbir Hasan, Adrien Lebre, Thomas Ledoux, Guillaume Le Louët, Jean-Marc Menaud, Jonathan Pastor, Rémy Pottier, Anthony Simonet, Mario Südholt.

Cloud applications and infrastructures

Complex event processing.   We presented this year the evolution of SensorScript towards a language for complex event processing dedicated to sensor networks. While the model mainly relies on previous works, we highlighted how the new language builds on the multitree in order to provide complex event processing mechanisms. We are able to balance the syntactic concision of the language with a real-time complex event processor for sensor networks. By providing flexible selections over the nodes, with the possibility to filter them on complex conditions, possibly over a time window, we offer a strong alternative to traditional SQL used in the literature. Moreover, SensorScript does not focus only on data access. In fact it provides the possibility to widen the scope of the methods accessible on nodes to other features than sensors monitoring, including but not limited to addressing actuators functions. Finally we showed that SensorScript is able to address examples proposed in the literature, with simpler results than SQL, while highlighting its limitations, especially on history management. [24]

Secure cloud storage.   The increasing number of cloud storage services like Dropbox or Google Drive allows users to store more and more data on the Internet. However, these services do not give users enough guarantees in protecting the privacy of their data. In order to limit the risk that the storage service scans user documents for commercial purposes, we propose a storage service that stores data on several cloud providers while preventing these providers to read user documents. TrustyDrive is a cloud storage service that protects the privacy of users by breaking user documents into blocks in order to spread them on several cloud providers. As cloud providers only own a part of the blocks and they do not know the block organization, they can not read user documents. Moreover, the storage service connects directly users and cloud providers without using a third-party as is generally the practice in cloud storage services. Consequently, users do not give critical information (security keys, passwords, etc.) to a third-party. [30]

Service-level agreement for the Cloud.

Quality-of-service and SLA guarantees are among the major challenges of cloud-based services. In [19], we first present a new cloud model called SLAaaS — SLA aware Service. SLAaaS considers QoS levels and SLA as first class citizens of cloud-based services. This model is orthogonal to other SaaS, PaaS, and IaaS cloud models, and may apply to any of them. More specifically, we make three contributions: (i) we provide a domain-specific language that allows to define SLA constraints in cloud services; (ii) we present a general control-theoretic approach for managing cloud service SLA; (iii) we apply our approach to MapReduce, locking, and e-commerce services.

Cloud Capacity Planning and Elasticity.

Capacity management is a process used to manage the capacity of IT services and the IT infrastructure. Its primary goal is to ensure that IT resources (services, infrastructure) are right-sized to meet current and future requirements in a cost-effective and timely manner. In [34], we present a comprehensive overview of capacity planning and management for cloud computing. First, we state the problem of capacity management in the context of cloud computing from the point of view of several service providers. Second, we provide a brief discussion about when capacity planning should take place. Finally, we survey a number of methods for capacity planning and management proposed by both people from industry and researchers.

In his PhD [12], Simon Dupont proposes to extend the concept of elasticity to higher layers of the cloud, and more precisely to the SaaS level. He presents the new concept of software elasticity by defining the ability of the software to adapt, ideally in an autonomous way, to cope with workload changes and/or limitations of IaaS elasticity. This brings the consideration of Cloud elasticity in a multi-layer way through the adaptation of all kind of Cloud resources (software, virtual machines, physical machines). In [23], we introduce ElaScript, a DSL that offers Cloud administrators a simple and concise way to define complex elasticity-based reconfiguration plans. ElaScript is capable of dealing with both infrastructure and software elasticities, independently or together, in a coordinated way. We validate our approach by first showing the interest to have a DSL offering multiple levels of control for Cloud elasticity, and then by showing its integration with a realistic well-known application benchmark deployed in OpenStack and Grid'5000 infrastructure testbed.

Infrastructure.

Academic and industry experts are now advocating for going from large-centralized Cloud Computing infrastructures to smaller ones massively distributed at the edge of the network (aka., Fog and Edge Computing solutions). Among the obstacles to the adoption of this model is the development of a convenient and powerful IaaS system capable of managing a significant number of remote data-centers in a unified way.

In 2016, we achieved three major results in this context.

The first result is related to the economical viability of Fog/Edge Computing infrastructures that is often debated w-r-t large cloud computing data centers operated by US giants such as Amazon, Google .... To answer such a question, we conducted a specific study that goes beyond the state of the art of the current cost model of Distributed Cloud infrastructures. First, we provided a classification of the different ways of deploying Distributed Cloud platforms. Then, we proposed a versatile cost model that can help new actors evaluate the viability of deploying a Fog/Edge Computing offer. We illustrated the relevance of our proposal by instantiating it over three use-cases and comparing them according to similar computation capabilities provided by the Amazon solution. Such a study clearly showed that deploying a Distributed Cloud infrastructure makes sense for telcos as well as new actors willing to enter the game [29].

The second result is related to the preliminary revisions we made in OpenStack. The OpenStack software suite has become the de facto open-source solution to operate, supervise and use a Cloud Computing infrastructure. Our objective is to study to what extent current OpenStack mechanisms can handle massively distributed cloud infrastructures and to propose revisions/extensions of internal mechanisms when appropriate. The work we conducted this year focused on the Nova service of OpenStack.More precisely, we modified the code base in order to use a distributed key/value store instead of the centralized SQL backend. We conducted several experiments that validate the correct behavior and gives performance trends of our prototype through an emulation of several data-centers using Grid’5000 testbed. In addition to paving the way to the first large-scale and Internet-wide IaaS manager, we expect this work will attract a community of specialists from both distributed system and network areas to address the Fog/Edge Computing challenges within the OpenStack ecosystem [36], [27]. These and additional corresponding results have been presented in a more detailed manner as part of Jonathan Pastor's PhD thesis [14].

The third result is related to the data management in Fog/Edge Computing infrastructures. Our ultimate goal is to propose an Amazon-S3 like system, i.e., a blob storage service, that can take into account Fog/Edge specifics. The study we achieved this year is preliminary. We first identified a list of properties a storage system should meet in this context. Second, we evaluated through performance analysis three “off-the-shelf” object store solutions, namely Rados, Cassandra and InterPlanetary File System (IPFS). In particular, we focused (i) on access times to push and get objects under different scenarios and (ii) on the amount of network traffic that is exchanged between the different sites during such operations. We also evaluated how the network latencies influence the access times and how the systems behave in case of network partitioning. Experiments have been conducted using the Yahoo Cloud System Benchmark (YCSB) on top of the Grid’5000 testbed. We showed that among the three tested solutions IPFS fills most of the criteria expected for a Fog/Edge computing infrastructure. [33], [32]

Renewable energy

With the emergence of the Future Internet and the dawning of new IT models such as cloud computing, the usage of data centers (DC), and consequently their power consumption, increase dramatically. Besides the ecological impact, the energy consumption is a predominant criterion for DC providers since it determines the daily cost of their infrastructure. As a consequence, power management becomes one of the main challenges for DC infrastructures and more generally for large-scale distributed systems. We have design the EpoCloud prototype, from hardware to middleware layers. This prototype aims at optimizing the energy consumption of mono-site Cloud DCs connected to the regular electrical grid and to renewable-energy sources. [17]

Green Energy awareness in SaaS Application.

With the proliferation of Cloud computing, data centers have to urgently face energy consumption issues. Although recent efforts such as the integration of renewable energy to data centers or energy efficient techniques in (virtual) machines contribute to the reduction of carbon footprint, creating green energy awareness around Interactive Cloud Applications by smartly using the presence of green energy has not been yet addressed. By awareness, we mean the inherited capability of SaaS applications to dynamically adapt with the availability of green energy and to reduce energy consumption while green energy is scarce or absent. In [25], we present two application controllers based on different metrics (e.g., availability of green energy, response time, user experience level). Based on extensive experiments with a real application benchmark and workloads in Grid'5000, results suggest that providers revenue can be increased as high as 64%, while 13% brown energy can be reduced without deprovisioning any physical or virtual resources at IaaS layer and 17 fold increment of performance can be guaranteed.