Modern web services are required to fulfill important non-functional criteria such as availability, responsiveness, scalability, and reliability, which are outlined in Service Level Agreements (SLAs). These agreements include Service Level Objectives (SLOs), which set performance benchmarks like up-time, latency, and throughput, crucial for maintaining consistent service quality. Failing to meet these SLOs can lead to penalties and harm to the provider's reputation. Additionally, service providers must avoid over-provisioning resources, as this can lead to excessive costs and inefficient use of resources. To address this, autoscaling mechanisms dynamically adjust the number of service replicas to match user demand, while scheduling policies establish a placement for each replica. However, traditional autoscaling solutions typically rely on low-level metrics (e.g., CPU or memory usage), making it difficult for providers to optimize both SLOs and infrastructure costs. Furthermore scheduling policies do not consider the run time resource contention between replicas that can have an impact on the application response time. This paper proposes an SLO-aware autoscaling methodology, along with load-aware scheduling and descheduling strategies, for containerized workloads in Kubernetes clusters, integrating SLOs in the container orchestration process. This approach overcomes the limitations of conventional orchestration policies by making more efficient decisions that balance service-level requirements with operational costs, offering a comprehensive solution for managing containerized applications and their infrastructure in Kubernetes environments. The results, obtained by evaluating a prototype of our system in a testbed environment, show significant advantages over the vanilla Kubernetes platform.

The orchestration of distributed microservices-based applications, particularly within geo-distributed Cloud-to-Edge environments, poses significant challenges. While Kubernetes stands as the predominant container orchestration standard in Cloud data centers, its static container scheduling approach presents limitations in deploying complex, distributed microservices-based applications across Edge environments. Presently, the scheduling process in Kubernetes fails to consider current infrastructure network conditions, resource usage, or runtime application statecrucial factors for mitigating the heterogeneous and dynamic nature of Cloud-to-Edge infrastructure and optimizing application response times. In this study, we propose an enhancement of the Kubernetes platform by implementing a load and network-aware microser-vices scheduling and orchestration strategy. The idea is to extend the Kubernetes control and scheduling logic with a dynamic orchestration strategy, continuously adapting application place-ment based on the real-time state of both the infrastructure and the application itself. We evaluate the efficacy of our approach by comparing it with the default Kubernetes orchestration and scheduling strategy.

Defining a scheduling and orchestration strategy for modern distributed microservices-based applications is a complex problem to deal with, especially if they are deployed on geo-distributed Cloud-to-Edge environments. Kubernetes is today the de-facto standard for container orchestration on Cloud data centers. However, its static container scheduling strategy is not suitable for the placement of complex and distributed microservices-based applications on Edge environments. Current infrastructure network conditions and resource availability neither run time application state are taken into account when scheduling microservices. To deal with these limitations in this work we present an extension of the Kubernetes platform in order to implement an effective application and infrastructure-aware container scheduling and orchestration strategy. In particular, we propose an extension of the default Kubernetes scheduler that considers application and infrastructure telemetry data when taking scheduling decisions. Furthermore, a descheduler component is also proposed that continuously tunes the application placement based on the ever changing application and infrastructure states. An evaluation of the proposed approach is presented by comparing it with the default Kubernetes scheduling strategy.

With the diffusion of Fog and Edge Computing paradigms, new application categories have emerged with specific quality of service requirements, in terms of communication latency and throughput. The placement of these applications on distributed Cloud-Edge environments is a challenging task, because of the continuously varying node-to-node network latency and bandwidth on Edge infrastructure. Although Kubernetes is the de-facto standard for container orchestration on Cloud data centers, its scheduling strategy is not suitable for the placement of time critical applications on Edge environments because it does not take into account current network conditions, neither communication requirements between microservices during its scheduling decisions. In this work we propose a network-aware scheduler plugin that extends the default Kubernetes scheduler, in order to deal with variable network conditions on cloud-edge clusters and run-time communication requirements of microservices. A custom descheduler is also proposed that periodically monitors run-time network state and traffic exchanged between microservices and evicts Pods from cluster nodes if better placement decisions can be done. An evaluation of our scheduling and descheduling strategies has been carried out on a sample microservices-based application deployed on a test bed environment.

Abstract The growth in the number and types of cloud-based services offered to IT customers is supported by the constant entry of new actors in the market and the consolidation of disruptive technologies such as AI, Big Data and Micro-services. From the customer’s perspective, in a market landscape where the cloud offer is highly diversified due to the presence of multiple competing service providers, picking the service that best accommodate their specific needs is a critical challenge. Once the choice is made, so called “cloud orchestration tools” (orchestrators) are required to take care of the customer application’s life-cycle. While big players offer their customers proprietary orchestrators, in the literature quite a number of open-source initiatives have launched multi-cloud orchestrators capable of transparently managing applications on top of the most representative cloud platforms. In this paper, we propose TORCH, a TOSCA-based framework for the deployment and orchestration of cloud applications, both classical and containerised, on multiple cloud providers. The framework assists the cloud customer in defining application requirements by using standard specification models. Unlike other multi-cloud orchestrators, adopts a strategy that separates the provisioning workflow from the actual invocation of proprietary cloud services API. The main benefit is the possibility to add support to any cloud platforms at a very low implementation cost. In the paper, we present a prototypal implementation of TORCH and showcase its interaction with two different container-based cluster platforms. Preliminary performance tests conducted on a small-scale test-bed confirm the potential of TORCH.

Although cloud computing has been around for over a decade now, many issues of the first hour still persist. Vendor lock-in, poor interoperability and portability hinder users from taking full advantage of main cloud features. On the industry side, the big players often adopt proprietary solutions to guarantee a seamless management and orchestration of cloud applications; on the research side, the TOSCA specification has emerged as the most authoritative effort for the interoperable description of cloud services. In this paper, we discuss the design and implementation of a TOSCA-based orchestration framework for the automated service provisioning. We also compare this approach to two open-source orchestration tools (Heat, Cloudify) with respect to the deployment of a two-tier application on an OpenStack environment. Test results show our method performs comparably to the aforementioned tools, while maintaining full compliance with TOSCA.

Hybrid and multi-cloud become prominent infrastructure strategy of enterprise. However, the complexity of such infrastructure is a considerable challenge. It is common to see that multi-cloud infrastructure is divided into several smaller units to facilitate the management. The division criteria are geolocation, cost, security, etc. Therefore, how to manage application deployment in such partitioned environment is an intriguing topic of multi-cloud management. We propose a policy-based deployment in multi-cloud infrastructure, which contains policy evaluation and TOSCA standard based orchestration. The system architecture is introduced and a case study with two empirical scenarios is discussed. The results indicate that the proposed policy-based deployment is useful in finding suitable resource and improving deployment efficiency.

Cloud computing is nowadays a consolidated paradigm that enables scalable access to computing resources and complex services. One of the greatest challenges Cloud providers have to deal with is to efficiently automate the service "provisioning" activities through Cloud orchestration techniques. By relying on TOSCA, a well-known standard specification for the interoperable description of Cloud services, we developed a fault-aware orchestrator capable of automating the workflow for service provisioning. The BPMN notation was used to define both the workflow and the data associated with workflow elements. To corroborate the proposal, a software prototype was developed and tested with a sample use case which is discussed in the paper.

The wide spread adoption of IoT technologies has resulted in generation of huge amount of data, or Big Data, which has to be collected, stored and processed through new techniques to produce value in the best possible way.Distributed computing frameworks such as Hadoop, based on the MapReduce paradigm, have been used to process such amounts of data by exploiting the computing power of many cluster nodes.Unfortunately, in many real big data applications the data to be processed reside in various computationally heterogeneous data centers distributed in different locations.In this context the Hadoop performance collapses dramatically.To face this issue, we developed a Hierarchical Hadoop Framework (H2F) capable of scheduling and distributing tasks among geographically distant clusters in a way that minimizes the overall jobs execution time.In this work the focus is put on the definition of a job scheduling system based on a one-point iterative search algorithm that increases the framework scalability while guaranteeing good job performance.

Effective big data analysis is one of the most notable research challenge of the latest few years. Hadoop, the most popular implementation of the MapReduce framework, has today become widespread used for processing large data sets using cloud resources. However, in many scenarios, data are geographically distributed over data centers and moving them to a single site for processing may result extremely expensive when not feasible at all. A key challenge for running applications in such a geographically distributed environment is how to efficiently schedule the computation over the different datacenters. In this work we present a job scheduler for a Hierarchical Hadoop Framework (H2F) that allows the management of multiple requests of job execution ensuring an efficient use of the available resources. Our experimental evaluations show that using H2F significantly improves processing time for geodistributed data sets with respect to a plain Hadoop system.

Big data analysis requires adequate infrastructure and programming paradigms capable of processing large amount of data. Hadoop, the most known open-source implementation of the MapReduce paradigm, is widely employed in big data analysis frameworks. However, in many recent application scenarios data are natively distributed over different geographic regions in data centers which are inter-connected through network links with very lower bandwidth than those of the computing environments where traditionally Hadoop deployments are supposed to work. In such a context, Hadoop applications perform very poorly. To cope with these issues, we developed a Hierarchical Hadoop Framework (H2F) specifically designed to work on geodistributed data. In this work, we compare the performance of H2F with that of a plain Hadoop implementation. First results show that for very large amount of data the H2F solution performs better than the Hadoop.

TEST 02 - Elsevier's Scopus, the largest abstract and citation database of peer-reviewed literature. Search and access research from the science, technology, medicine, social sciences and arts and humanities fields.

In recent years we have been witnessing significant advances in the medical imaging techniques. With respect to the past, current imaging processes enable physicians to reach higher precision level when it comes to diagnosing patients' diseases. Such a precision is supported by sophisticated imaging machinery, whose cost increases with the technological advances. Also, most of imaging scans expose patients to ionizing radiation which in certain cases may seriously harm their health. In this panorama, ordering inappropriate imaging examinations has a great impact on both the economy of a nation and on its citizens' health. Many governments have published guidelines which physicians should follow when deciding which imaging technique is to be prescribed to best answer to a clinical question. The COLLABORADI project is a recently concluded research project which studied the phenomenon of inappropriate imaging prescriptions in Sicily (Italy), and proposed the design and implementation of a decision support system to help physicians to set up the most appropriate diagnostic route for their patients. This paper discusses the work carried out within the project.

In the past two decades there has been a growing interest over the definition of new distributed computational paradigms capable to serve the need of manipulating and analyzing huge amounts of data. Among the others, the MapReduce outstands for popularity. Its open-source implementation Hadoop is widely used in academic environments and is also greatly supported by huge IT players. There are many application scenarios where the data to be manipulated resides on data centers which are heterogeneous in term of computing capacity and are geographically distant from each other's. Unfortunately, in this contexts Hadoop performs very poorly. In this paper we propose to leverage on a hierarchical computing framework to boost the Hadoop performance in geo-distributed computing environments. The framework we propose drains fresh information from the distributed computing context and exploits it to carry out a smart job scheduling strategy. In this work, the focus is put on the study and definition of the application profile of the jobs. We implemented a software prototype of the proposed hierarchical Hadoop framework. Tests run on the prototype proved the capability of the job scheduling system to compute the job's execution path and estimate its completion time.

The fourth industrial revolution related to the Internet of Things (IoT) is leading to the emergence of future smart cities and cyber-physical systems with a rapidly increasing of smart connected objects (such as personal devices, sensors, actuators). Therefore, companies such as Telcos, Cloud and Service Providers have to analyze a huge amount of data that probably lead to the reduction of available bandwidth and subsequent increase of network latency. In this context, Fog Computing and Mobile Edge Computing (MEC) can be suitable paradigms to solve this kind of problems. The exploitation of the heterogeneous environment allows a better performance for mission-critical and time-critical applications. In this paper we present a scenario that exploits the MECs in order to detect abnormal or critical events such as terrorist threats, natural and man-made disasters. The proposed solution allows cooperation among the Base Transceiver Stations (BTS) to rapidly notify the users which are close to the critical area.

TEST 02 - Elsevier's Scopus, the largest abstract and citation database of peer-reviewed literature. Search and access research from the science, technology, medicine, social sciences and arts and humanities fields.

Currently in the cloud market resources are mainly allocated according to the fixed-price, direct selling model. Market principles such as the supply-demand rate are not taken into consideration by cloud providers. In the literature several market-based resource allocation models and algorithms have been proposed, showing that dynamic pricing models might result more convenient for both the providers and the consumers of cloud resources. In this paper, we discuss about the use of alternative auction-based mechanisms to sell the “residual” computing capacity, i.e., the capacity which the provider is not able to allocate through direct-selling. The design of a procurement market for computing resources is proposed: in such a market we devised an adaptive bidding strategy that suggests to the provider the right actions useful to attain its business objective. The resource overbooking mechanism is also proposed to overcome the problem of resource underutilization. A software simulator was implemented with the aim of testing and validating the proposed mechanisms. Results show that, by fine-tuning their own strategy, providers manage to pursue specific objectives.

TEST 02 - Elsevier's Scopus, the largest abstract and citation database of peer-reviewed literature. Search and access research from the science, technology, medicine, social sciences and arts and humanities fields.

Nowadays sensors and sensor networks are deployed in many application domains to accomplish well defined, field-oriented monitoring tasks. Very often, sensors which have been deployed to monitor specific contexts are required to interoperate to third-party sensor systems to serve more complex and demanding monitoring tasks. The heterogeneity of sensor platforms offered by vendors raises the need for a common framework where sensors can be regarded as vendoragnostic services with well defined and open functionality. To this end, this work proposes a semantic approach to foster the characterization of heterogeneous sensors. Also, an effort is made to semantically characterize the geographic context where sensors may be deployed, in a way that in the resulting semantic framework functional and geographic properties of sensors may be combined together. The exploitability of such an idea is then investigated through a practical use case in the domain of civil protection. The implemented scenario shows that the combination of sensors' functional and geographic properties is useful to the domain's monitoring purpose.

Cloud computing has reached a high level of acceptance, both in academia and in industry. The maturity of the technology, along with the considerable business opportunity that has been promised, is the main responsible for its success. Nevertheless, today only few, very big players dominate the commercial panorama and take over market shares. From that position, they impose rigid pricing policies and quite inflexible negotiation schemes. When the ongoing cloud standardization process will complete, and thus full interoperability among clouds will be accomplished, new players will come into play. A real competition among cloud providers will then start based on key factors like the capability of providing flexible services tailored to specific, fine-grained customers' requirements. In this market scenario a mechanism must be devised to support the matchmaking between what providers offer and what customers demand. In this work we define a semantic model to help customers and providers to characterize their demands/offers, and propose the use of semantic tools to perform the matchmaking in such a way to maximize both the provider's and the customer's satisfaction.

Cloud computing poses several new security and privacy challenges, mainly related to resource sharing, interoperability and dinamicity among different providers. Although policy specification languages address some of these challenges,many issues still have to be faced with. Policy matching is today performed by way of syntactical approaches, which may limit the selection of suitable services on the one hand, and the flexibility and the dinamicity of the matching process on the other one. In this work we propose a semantic approach that,by means of semantic annotations to WS-Policy documents,allows for an improved matching of security requirements and capabilities based on their actual meaning. The proposed approach has been validated through a case study that shows how a pure syntactic-based mechanism of WS-Policy would have failed in matching two actually compatible policies.

In SOA environments, as the number of published Web Services grows, the provision of a robust, scalable and efficient discovery service is still an unresolved issue. In this paper we present a P2P-based infrastructure that leverages the semantic technologies to support a scalable and accurate service discovery process. The key concept of the presented idea is the creation of an overlay network organized in several semantic groups of peers, each specialized in answering queries pertaining to specific applicative domains. Groups are formed by clustering together peers offering services that are semantically related. A semantic query routing mechanism ensures the delivery of queries to the groups that are semantically "worth" to respond. In this paper the architecture details of the proposed solution is presented. A system prototype has also been implemented and validated through a case study deployed on the PlanetLab testbed.

Semantic web services have recently appeared as an enabling technology for the challenging task of automated service discovery and composition. In addition, the growing number of web services demands for distributed discovery mechanisms which do not suffer of the scalability problems of the centralized ones. A framework based on the P2P paradigm to enhance the service discovery process has been designed: it is mainly based on the grouping of peers which are semantically "close" to each other. In this paper, we focus on the semantic characterization of a peer, which is built by taking into account the set of services it provides. In order to reach an high level of service integration, our framework builds an abstract "expertise" of the peer, which is independent of the specific semantic language used to annotate the service.

A large number of heterogeneous and computing devices, such as PCs, PDAs, and cell phones, nowadays are used to access the same information. Currently, designers designing such multi-device user interfaces have to design a UI separately for each device, which is a time consuming and error prone activity. This paper discusses our approach to the multi-device interface development. In particular we describe how abstract UI descriptions and task model management systems can be combined to develop adaptive UIs for a wide range of devices. The designed software framework allows generating the concrete user interface at runtime, by adapting it to the client's execution environment. As shown in the example application, three different environments have been the target of our implementation work: standard PCs, PDAs and mobile phones equipped with Java micro edition

The availability of powerful personal computers and high bandwidth home connectivity is changing the way users cooperate and exchange information on the network. Peer to peer computing is becoming a common paradigm for many distributed applications, allowing for extensive resource sharing and direct communication among peers. Recently, the wide spreading of new wireless communication technologies and personal computing devices, enables the adoption of the P2P paradigm also in mobile environments. However, several questions arise when trying to develop such systems: existing solutions do not address all the requirements of these environments. In this paper we present a framework to develop mobile applications which exploits the P2P paradigm: it is designed to work on J2ME enabled mobile devices, interacting with each other in an ad-hoc fashion. The designed framework is also interoperable with JXTA, a well-known P2P open platform.

Peer-to-peer (P2P) technology has arisen as a simple and user-friendly mechanism not relying on a central server. In this paper an attempt to exploit this emerging technology in the area of centralization in a local region with flexibility of P2P mechanism on a broader level. The implementation of the proposed approach, which has been built on top of the MAP agent system, is compared with more traditional techniques for agent location, based on a naming service. Comparisons between the two approaches are presented together with some hints helping to identify the best approach in each considered scenario.

The combination of personal computing devices and wireless ad-hoc networks allows the concept of mobile ad-hoc information system, consisting of a highly dynamic, decentralized and self-organizing network of autonomous and mobile devices that interact as peers [M. Hannicainen et al. (2002), L. Kleinrock, (2000)]. Here we define a middleware layer, named Expeerience. We have decided to adopt as a basic software framework, an emerging P2P open technology such as JXTA. The second issue we address in this work is the integration of some code mobility support in the developed middleware to allow the distribution and execution of services on peers that originally do not own the service code. This has been introduced in our platform by adding a mobile code service, which enables to download and install new services dynamically only if necessary.

Since its appearance, mobile agent technology has been considered as a new and flexible paradigm for distributed application development. However because of code mobility, this paradigm produces new threats to security too, which are of central importance when proposing this technology for commercial applications. We show the security services which have been specifically designed and developed for the MAP agent system. Early performance measurements show how the overhead needed by the security mechanisms introduced can be acceptable if compared with the functionalities provided.

The third-generation mobile networks will provide new and advanced services for mobile users: in this context, the virtual home environment (VHE), currently being widely investigated, is a concept which will enable users to access and personalize their subscribed services whatever the terminal they use and whatever the underlying network used. Adaptation is a key feature for these architectures: services should be designed and developed only once, independently of the actual user terminal and its capabilities. We show how an adaptation framework could benefit of code mobility, especially for services providing multimedia content. The base architecture is shown, and specific scenarios are discussed. The work in this paper has been done in the context of the European IST project VESPER.

Agent programming technology has emerged as a flexible and complementary way to manage resources of distributed systems due to the increased flexibility in adapting to the dynamically changing requirements of such systems. A very promising application of this technology is related to the control of forthcoming networking systems which will represent a competitive marketplace with a multitude of vendors, operators and customers. Thus, new reference models have to be investigated in order to better satisfy users' requirements in a framework where resource allocation is provided under the control of different and often competing stakeholders (users, network providers, service providers, etc.). We believe that autonomy is one of the features that will characterize the behavior of agents in such environment: autonomous choices will be taken as the result of coordination among different cooperating software entities. Following this direction, we describe the efficient integration and adoption of mobile agents and genetic algorithms in the implementation of a valuable strategy for the development of effective market based routes for brokering purposes in the future multioperator network marketplace. The proposed genetic algorithm provides a kind of stochastic algorithm searching process in order to identify optimal resource allocation strategies. The agent-based network management approach represents an underlying framework and structure for the multioperator network model, and can be used to facilitate the collection and dissemination of the required management data, as well as the efficient and distributed operation of the algorithm. We also present some numerical results to assess the performance and operation effectiveness of our approach, by applying it in some test case scenarios.

Grid systems are becoming the underlying infrastructure for many high-performance distributed scientific applications. Resources in these systems have heterogeneous features, are connected by potentially unreliable networks, and are often under different administrative management domains. These remarks show how a new generation of techniques, mechanisms and tools need to be considered that can cope with the complexity of such systems and provide the performance suitable for scientific computing applications. In this paper an architecture for monitoring services in Grid environments based on mobile agent technology is presented. One of the most interesting aspects of the proposed approach is the possibility of customizing the parameters to be monitored. After describing the general framework where agents will execute, an application for improving the capabilities of traditional network monitoring paradigms is presented.

An agent based approach is investigated to build a framework where resource allocation is provided under the control of different and often competing stake-holders (users, network providers, service providers, etc.). This paper also describes the efficient integration and adoption of mobile agents and genetic algorithms in the implementation of an effective strategy for the development of effective market based routes for brokering purposes in the future multi-operator network marketplace. The agent-based network management approach represents an underlying framework and structure for the multi-operator network model, and can be used to collect all the required management data. The proposed genetic algorithm provides a kind of stochastic algorithm searching process in order to identify optimal resource allocation strategies.

The authors present the design and the implementation of a distributed media service provider for the Internet. The overall architecture consists of several remote video servers whose resources are merged into a logically unique and more powerful service provider. The basic abstraction layer needed to achieve inter-server resource coordination and user-server interoperability has been realized through the use of OMG's CORBA and SUN's Java facilities. The current implementation of the video servers is based on the DAVID video server developed at the University of Catania (http://sun195.iit.unict.it/david).

A previous development in the field of telecommunication networks has enabled the evolution of distributed systems, and has favoured the development of a new generation of applications based on the use of mobile agents. In fact, during the last few years, several agent platforms have been created; however the architectures and implementation of such platforms are very different, and this makes their interaction difficult. In order to favour the interaction among differently implemented platforms, the Object Management Group (OMG) has proposed the Mobile Agent System Interoperability Facility (MASIF) standard, which includes some definitions and interfaces which all platforms have to comply with. We present the MAP platform for developing and managing mobile agents; it complies with MASIF in order to communicate and interact with similar platforms provided that they comply with the mentioned standard. Besides, we present a mechanism, which has been implemented in the MAP, which extends MASIF's functionalities so to enable the interaction among agents and agent systems operating in different regions.

Network programmability seems to be a promising solution to network management and quality of service (QoS) control. Software mobile-agents technology is boosting the evolution toward application-level control of network functionalities. Code may be deployed in the network dynamically and on demand for the benefit of applications or application classes. Agents support a dynamic distribution of control and management functions across networks, thus increasing flexibility and efficiency. We propose to use mobile-agent technology to overcome some of the problems inherent in current Internet technology. We focus our attention to QoS monitoring, being locally significant in network subdomains, and realize a QoS management strategy in response to variations of user, customer of application requirements, and of the network state. We describe our experience and the results obtained from our testbed, where software agents are instantiated, executed, migrated, and suspended in order to implement flexible QoS management in IP networks.

Originally, the Internet delivered best-effort service quality with respect to end-to-end delay. Recently, extensions such as RSVP have been proposed to provide guaranteed real-time services as well. Unfortunately, network resources, such as routers, do not yet fully, support RSVP reservation protocols so that guarantees cannot truly be given. We suggest to follow the paradigm of open programmable networks for a more complete QoS provisioning. Reservation gaps or tunnels are dynamically closed by means of a software agent approach that is flexibly deployed for an application oriented QoS support. Agents are dynamically located to such tunnels in order to monitor the tunnels, to provide feedback information in case of QoS violations, and to decide on possible compensating measures to be taken.

As distributed multimedia applications become more widely diffused, systems for quality of service QoS management are increasingly essential. In this paper we present an architecture for distributed QoS management based on software agents. We discuss the advantages of using agent technology in a problem inherently distributed and complex such as QoS management. Besides introducing several negotiation and coordination techniques among agents as fundamental properties for QoS management, it is argued in favor of decision making based on analytical and simulative modeling techniques. Furthermore, design and implementation issues are discussed in detail and our Java-based platform for agent-based QoS management is introduced.