In recent years, mobile cloud computing has attracted researchers and industries' interest since it maximizes gains with mobile computing in different areas, such as healthcare, sports, and business. However, the inherent challenges associated with mobile computing (e.g., battery provisioning and wireless network instability) may limit mobile cloud achievements. This paper provides an availability and reliability study of mobile clouds and evaluates distinct architectures and strategies. Our evaluation is based on hierarchical heterogeneous models and focuses on steady-state availability, reliability, annual downtime, and costs of implementation and provisioning of the cloud infrastructure. The results point out that service downtime may be significantly decreased by adopting actions such as spare batteries for mobile devices and the adoption of distinct redundancy strategies. We also propose a way to support decision-making considering private and public clouds by comparing some architectural choices while considering factors such as budget expenses for hardware purchases and power consumption. After 36 months, the private cloud has a cumulative cost that is smaller than the cost of a public cloud service.

Multi-access Edge Computing (MEC) is a key enabler of 5 G networks, supporting handling latency, bandwidth, and connectivity requirements. However, integrating MEC with Software-Defined Networking (SDN) introduces significant challenges in resource management, Quality of Service (QoS) assurance, and scalability. Existing performance evaluation methods do not meet the expectations of addressing the complexity and reliability needed for SDN-based MEC systems. This paper proposes a stochastic modeling framework based on Stochastic Petri Nets (SPN) to assess the performance of MEC-SDN architectures. The model captures workload dynamics, system utilization, and failure conditions, providing a comprehensive and scalable performance evaluation tool. The method is validated through a real-world Vehicle-to-Infrastructure (V2I) scenario deployed on a cluster of single-board computers. Experimental results demonstrate up to 99.85% system utilization, low bandwidth consumption, and model validation within a 95% confidence interval. These outcomes confirm the model's effectiveness in evaluating SDN-enabled MEC deployments' resource allocation and performance.

Vehicular Cloud Computing (VCC) introduces the challenge of resource volatility, as vehicles unpredictably depart, impacting system stability and performance. To address this, we present a Stochastic Petri Net (SPN) model with a container allocation strategy to evaluate Static Vehicular Clouds (SVCs) in a car rental setting. Two key performance metrics are analyzed: Queue Discard Probability (QDP) and Utilization of Available Resources (UAR). A sensitivity analysis reveals that mean rental time (MRT), mean return time (MRRT), number of tasks per vehicle (NTV), and mean service time (MTST) are the primary factors influencing system behavior. Two case studies, varying NTV and MTST, demonstrate their impact on QDP and UAR. These findings offer valuable guidance to infrastructure planners aiming to optimize resource allocation, minimize task loss, and improve SVC efficiency.

Reliable signal reception and attenuation control are critical to improving the performance of low-power wide area networks (LPWANs), especially in hybrid systems integrating LoRa and satellite communications. This study analyzes the effect of Graham’s algorithm in hybrid LoRa-satellite networks using NEO-M8N GPS-enabled TTGO nodes deployed in mesh in hindered environments. Implementation of the algorithm optimized inter-node connectivity, reduced signal attenuation, and improved satellite availability. A superior packet delivery ratio (PDR) was observed with the Graham algorithm. In addition, power consumption per minute was stabilized and battery life was extended. Through mathematical modeling and experimental validation, it is confirmed that Graham improves the energy efficiency and robustness of the network, enhancing the performance of hybrid LPWAN systems.

This paper presents a comprehensive Stochastic Petri Net (SPN) model for evaluating energy availability in multi-microgrid systems, integrating photovoltaic (PV) generation, battery storage, and distribution networks. The proposed framework addresses the inherent uncertainties of renewable energy sources (RES), dynamic load demands, and operational transitions between grid-connected and islanded modes. By leveraging SPNs, the model captures the probabilistic behavior of microgrid components, enabling precise assessment of steady-state and instantaneous availability metrics. A Reliability Block Diagram (RBD) further extends the analysis to cooperative microgrid networks, quantifying the collective ability to meet contractual energy commitments under shared surplus scenarios. Simulation results, validated using the CIGRE LV benchmark network, demonstrate the model’s efficacy in identifying reliability bottlenecks, optimizing maintenance strategies, and enhancing system resilience. The study highlights the SPN’s versatility in modeling complex interactions across residential, industrial, and commercial microgrids, providing actionable insights for sustainable energy system design.

As urban centers increasingly grapple with the challenges of population growth, traffic congestion, and environmental degradation, the need for sustainable mobility solutions has become imperative. This study presents a performance evaluation of a public bicycle-sharing system through the application of Stochastic Petri Nets (SPNs), which are employed to model and simulate user behavior under dynamic conditions. Drawing on recent advances in urban mobility and shared transportation research, the proposed stochastic framework effectively captures key operational metrics, such as bicycles in transit, station availability, waiting probability, system utilization, and the occupancy of docking stations. A case study analyzes three high-demand stations across varying demand scenarios (baseline, 25%, and +25%). Findings reveal that, while the system performs adequately under current demand levels, it becomes vulnerable to saturation with increased usage and demonstrates operational inefficiencies when demand decreases. The model offers valuable insights to inform operational strategies, guide infrastructure development, and support public policy aimed at fostering integrated and sustainable urban mobility.

The evolution of mobile technology from the fifth-generation (5G) radio access has amplified challenges. As new requirements emerge, there is a need for enhanced computational capacity for planning, operation, and availability posed by 5G. Multi-access Edge Computing (MEC) and Software-defined networks (SDN) are fundamental to addressing capacity planning and availability. This paper presents a hierarchical modeling approach using reliability block diagrams (RBD) and continuous-time Markov chains (CTMC) to estimate the MEC SDN-based availability and capacity-oriented availability (COA). The proposed models allow analytical evaluations for calculating dependability metrics, showing an availability increase from 97.33% to 99.58%, with only two clustered server nodes, effectively reducing annual downtime from 357.32 hours to 0.65 hours.

In this paper, we present a hierarchical modeling strategy that combines stochastic Petri nets (SPNs) with an iterative algorithm to evaluate performance in containerized microservices. Our approach models both synchronous and asynchronous calls, bounded queues, and network constraints, and then dynamically adjusts the number of replicas to keep the discard probability within predefined thresholds. By analyzing each microservice in a modular fashion, we avoid exploding the state space and achieve accurate predictions for different load and bandwidth scenarios. In experiments with 21 containerized microservices under three bandwidth constraints (10, 40 and 100 MB/s), our method correctly predicted throughput, container usage and bandwidth consumption within 95% confidence intervals, confirming its effectiveness for resource allocation decisions in microservice platforms.

This paper introduces a stochastic model to characterize battery degradation using Stochastic Petri Nets (SPN). The model analyzes battery degradation under varying discharge currents, architectural configurations, state of health (SoH) thresholds, K-out-of-N (KooN) architectures, and cell imbalances. It assesses reliability based on individual cell data, distinguishing it from traditional methods by accounting for cell failure influence. A NASA dataset was used to study battery degradation trends under various imbalance levels. Probability distributions for the SoH decay process were incorporated into the models, and simulations were conducted to observe battery behavior across cell imbalance ranges from 5% to 80% of the SoH amplitude. Results confirm that the model effectively reflects degradation behavior related to battery reliability. Findings indicate the parallel-series (PS) configuration achieves up to 18.5% higher mean cycles to failure (MCTF) than series-parallel (SP) under significant imbalance, with modest advantages (around 3%) under balanced conditions. This shows the model accurately assesses reliability and can be adapted for battery packs.

This paper presents a stochastic performance model for evaluating battery swapping station (BSS) systems for electric vehicles (EV). The approach uses Stochastic Petri Nets (SPN) and discrete event simulation to model interactions among vehicles, battery swapping infrastructure, and demand distribution. A modular framework simulates a real-world case study in Recife, Brazil. The Average Probability of Not Finding a Battery (APNFB) is evaluated under various operational scenarios. A sensitivity analysis highlights key parameters affecting system performance: battery discharge time, charge time, and the number of motorcycles. These parameters significantly impact battery availability and reliability, emphasizing the need for informed infrastructure planning and design. The study illustrates the value of stochastic modeling in understanding BSS system performance under uncertainty, due to the aleatory behavior of the system, and supports strategic decision-making for future deployments.

Moving Target Defense (MTD) consists of applying dynamic reconfiguration in the defensive side of the attack-defense cybersecurity game. Virtual Machine (VM) migration could be used as MTD against specific host-based attacks in the cloud computing environment by remapping the distribution of VMs in the existing physical hosts. This way, when the attacker’s VM is moved to a different machine, the attack has to be restarted. However, one significant gap here is how to select a proper VM migration-based MTD schedule to reach the desired levels of system protection. This paper develops a Stochastic Petri Net (SPN) model to address this issue. The model leverages empirical knowledge about the dynamics of the attack defense in a VM migration-enabled setup. First, we present the results of an experimental campaign to acquire knowledge about the system’s behavior. The experiments provide insights for the model design. Then, based on the model, we propose a tool named PyMTDEvaluator , which provides a graphical interface that serves as a wrapper for the simulation environment of the model. Finally, we exercise the tool using Multi-Criteria Decision-Making methods to aid the MTD policy selection. Hopefully, our results and methods will be helpful for system managers and cybersecurity professionals. • Comprehensive experimental results of VM migration as MTD against Memory DoS. • Availability and security Stochastic Petri Net of a system with VM migration as MTD. • An open-source simulation tool providing a graphical interface for the model. • Multi-Criteria Decision-Making to support MTD selection based on user preferences.

Satellite communications have become an everyday aspect of our daily lives, and the current focus on satellite telecommunications is expected to continue to grow and gain greater prominence. On the other hand, implementing scheduled and unscheduled maintenance practices is essential to improving the availability and reliability of satellite systems. This paper presents a stochastic availability model designed through a Stochastic Petri Net (SPN) that considers the aging of satellites, offering support for infrastructure planning by integrating both scheduled and unscheduled maintenance considerations. Additionally, this research includes two case studies. The first case study explores an availability model that ignores the influence of aging, establishing a baseline of availability and second with aging. This baseline will be compared with the results obtained from the proposed SPN model, which incorporates the effects of aging. In this way, the behavior of both non-aging and aging variables can be analyzed to make contributions through the SPN that helps understand the availability of satellite systems.

Internet of Things (IoT) applications equip rural producers with decision support tools and automated solutions that boost agribusiness productivity, quality, and profit. However, most poultry farmers still use conventional methods of operation in which human workers carry out all routines for monitoring and controlling their farms at the expense of greater productivity. One of these human activities is manual weighing, which can be replaced by nonintrusive methods such as computational vision applications that estimate live poultry's weight using video cameras. Since Internet of Things (IoT) devices may have low computing power limiting the ability to process the data locally, they can transfer it to a fog or cloud data center, where they are processed. This article aims to conduct a dependability study of a poultry house automated with a computer vision-based system for estimating poultry weight considering hierarchical models (e.g., Markov chain, reliability block diagram, and closed-form equation) to represent the whole system and obtain steady-state availability and annual downtime. In addition, our purpose is to consider and compare different architectural solutions, such as edge and fog computing-based solutions. The proposed solution verified that a cloud-based application with no redundancy has a downtime of 34.14% and 9.176% hours when considering a hot-standby redundancy strategy in the office node of a cloud solution.

In recent years, machine learning models have used data from Hardware Performance Counters (HPCs) proposed to combat Distributed Denial of Service (DDoS). Several models use feature selection (FS) methods to deal with these datasets’ high dimensionality, seeking to develop efficient real-time detection systems. In this work, we propose a methodology to evaluate the behavior of 24 HPCs whose data is subjected to five different DDoS attacks. We adopt the FS methods Filter, Wrapper, Hybrid, and Embedded as a selection method. The results showed that three main HPCs are the most impacted and may be used to detect the attacks, keeping the results of the classification algorithms used in validation above 98%, which means an 87.5% reduction in the dimensionality of the dataset. The research results include that the Filter Approach demonstrated excellent stability in indicating the best HPCs for DDoS detection performance. In contrast, hybrid methods showed more unstable results, expanding the possibilities for future studies on the explainability of these methods. Finally, selecting essential features to detect various attack types can reduce the HPC data collection time or even discard multiplexing techniques, allowing the use of the proposed technique for real-time attack detection environments and contributing to the security of devices with fewer HPCs available, such as embedded and IoT devices.

Augmented Reality (AR) Systems have been largely adopted as a useful technology. They are excellent candidates for AR application execution platforms due to their recent popularity. Nonetheless, the limited processing power and battery autonomy restrict the possibilities of more extensive use of smartphones as a broadly adopted AR platform. This work explores the use of Edge Computing and Code Offloading in a collaborative AR system. We present the proposed application architecture, which includes several interconnected components divided into three primary blocks. The benefits of AR are discussed, including its use in education, medical training, and industry. The central objective of this work is to provide an in-depth and structured analysis of the proposed architecture, paving the way for extensive studies that address various dependability issues. We implemented an AR system to validate the proposed architecture. We collected the data related to the execution times of the algorithms allocated to the Edge, where it was possible to characterize a probability distribution suitable for characterizing these times. The proposed architecture was validated, and the results obtained open possibilities for future developments and better adequation of the applications implemented according to the proposed architecture.

The hybrid cloud inherits the best aspects of both the public and private clouds. One such benefit is maintaining control of data processing in a private cloud whilst having nearly elastic resource availability in the public cloud. However, the public and private cloud combination introduces complexities such as incompatible security and control mechanisms, among others. The result is a reduced consistency of data processing and control policies in the different cloud deployment models. Cloud load-balancing is one control mechanism for routing applications to appropriate processing servers in compliance with the policies of the adopting organization. This article presents a process-mining influenced load-balancer for routing applications and data according to dynamically defined business rules. We use a high-level Colored Petri Net (CPN) to derive a model for the process mining-influenced load-balancer and validate the model employing live data from a selected hospital.

The use of Virtual Machine (VM) migration as support for software rejuvenation was introduced more than a decade ago. Since then, several works have validated this approach from experimental and theoretical perspectives. Recently, some works shed light on the possibility of using the same technique as Moving Target Defense (MTD). However, to date, no work evaluated the availability and security levels while applying VM migration for both rejuvenation and MTD (multipurpose VM migration). In this paper, we conduct a comprehensive evaluation using Stochastic Petri Net (SPN) models to tackle this challenge. The evaluation covers the steady-state system availability, expected MTD protection, and related metrics of a system under time-based multipurpose VM migration. Results show that the availability and security improvement due to VM migration deployment surpasses 50% in the best scenarios. However, there is a trade-off between availability and security metrics, meaning that improving one implies compromising the other.

This paper presents PyMTDEvaluator, a tool for evaluating the effectiveness of time-based Moving Target Defense (MTD) against availability attacks (e.g., Denial of Service - DoS, resource starvation attacks). PyMTDEvaluator is based on simulation runs of an extended deterministic Stochastic Petri Net (SPN) and offers a user-friendly interface where it is possible to analyze and compare MTD policies with different parameters. The SPN design relies on knowledge obtained from empirical observation. PyMTDEvaluator provides results such as probability of attack success, availability, and system capacity to support MTD design decision making. The tool allows analyzing and comparing several scenarios in the same evaluation, thus enabling the study of the pros and cons of different MTD deployment alternatives. PyMTDEvaluator aims to be part of the toolset for MTD policies design. It is also valuable for sensitivity analysis of MTD-enabled system parameters.

The demand for various multimedia services has influenced the convergence of existing communication networks, allowing the integration of multiple traffic types through a single routable protocol (IP). Thus, we can adopt different strategies to meet quality requirements, ensure uninterrupted operations, high reliability, and lower maintenance costs. The model-based assessment enables the analysis of systems dependability without interference in the existing system and makes it possible to evaluate complex scenarios whose actual implementation may be impracticable due to costs. Evaluating a system in operation may not be trivial, or it can hinder the continuity of services. In this sense, we evaluated converged network infrastructures considering dependability metrics such as availability, reliability, mean time to failure, and downtime. We use sensitivity analysis to identify the components that can be essential on availability impact. The results allow us to identify the bottlenecks of the represented architectures.

Moving Target Defense (MTD) is a defensive mechanism based on dynamic system reconfiguration to prevent or thwart cyberattacks. In the last years, considerable progress has been made regarding MTD approaches for virtualized environments, and Virtual Machine (VM) migration is the core of most of these approaches. However, VM migration produces system downtime, meaning that each MTD reconfiguration affects system availability. Therefore, a method for a combined evaluation of availability and security is of utmost importance for VM migration-based MTD design. In this paper, we propose a Stochastic Reward Net (SRN) for the probability of attack success and availability evaluation of an MTD based on VM migration scheduling. We study the MTD system under different conditions regarding 1) VM migration scheduling, 2) VM migration failure probability, and 3) attack success rate. Our results highlight the tradeoff between availability and security when applying MTD based on VM migration. The approach and results may provide inputs for designing and evaluating MTD policies based on VM migration.

Cloud-based architectures have grown in recent years, especially the interest in container-based solutions have sharply increased by enterprises worldwide. Containers are a form of lightweight virtualization that can be used to provide cloud services. Adopting this kind of technology in a bare- metal context is becoming strong because they can offer many benefits, like performance efficiency and costs reduction. Docker is a widespread platform for the creation and management of containers. As in any computational cloud service, Docker environments must deal with the intensive workload and may have a long-term life cycle, which might trigger some problems that compromise the system dependability. The software aging phenomenon is one of these likely problems. It is a process of cumulative errors or system misbehavior that leads to application failures and performance degradation throughout its runtime. This paper aims to monitor and evaluate software aging effects on the Docker platform in a cloud computing environment. We conducted two experimental studies with automated workloads to simulate containers' life cycle and the intensive use of Docker features, while the system was monitored. The results show high resource consumption by the operating system's network utility, in addition to memory fragmentation in the sub-processes of the Docker platform. Trends of increasing resident memory consumption were also observed in one of these scenarios.

Internet of Things (IoT) is a network of physical objects equipped with embedded technology, sensors, and connection to the network (e.g. vehicles, buildings, and others). These physical objects are able to collect data that will be used by various applications. Currently, there are already various applications in IoT such as smart agriculture and smart parking. These applications are composed by a set of heterogeneous components, so tasks such as resource management and maintenance of these systems become complex. Therefore, this paper presents a modeling strategy based on hierarchical models using Stochastic Petri Net (SPN) and Reliability Block Diagram (RBD). In order to evaluate the dependability in an IoT system, a case study was carried out based on a specific scenario considering a smart building. The proposed models enable to estimate measures such as steady-state availability and annual downtime.

The ossification of Internet's architecture is widely recognized as an important problem and network virtualization is a remarkable approach to deal with it. In this context, the optimal allocation of substrate resources is a NP-hard problem and there are many works proposing efficient mapping of virtual networks (VN). Evaluation and analysis of throughput on VNs are also important, as it directly impacts the quality of service of virtualized networks. This paper proposes a technique based on stochastic Petri nets (SPN) and reliability block diagrams (RBD) to assess availability-oriented throughput in virtualized network environments, using the availability of each physical network component to estimate VN throughput. Experimental results demonstrate throughput degradation when the distance of physical nodes increases. This work also presents a fault tolerance technique that mitigates such a degradation.

Relational databases are the most popular database system worldwide. The occurrence of failures in these systems may produce severe consequences for the business, such as data loss, customer dissatisfaction, and subsequent revenue loss. Consequently, many organizations have adopted disaster recovery (DR) solutions as an attempt to prevent data loss and ensure business continuity. Data replication for databases is one of the most used DR solution employed to guarantee data safety and availability. However, the analysis regarding DR aspects has been less explored. Therefore, in this paper, we present an integrated model-experiment approach to evaluate replication mechanisms in relational databases for DR purposes. We performed experiments in a geo-distributed cloud environment and developed analytic models to evaluate DR key-metrics such as availability, downtime, Recovery Time Objective (RTO), and Recovery Point Objective (RPO). The results revealed that the adoption of replication mechanisms could increase the system's availability significantly. It also revealed that the replication mechanisms can guarantee RPO and RTO within seconds.

Demand for performance, availability, and reliability in computational systems has increased lately. Improving these aspects is an important research challenge due to the applications and users diversity. In this paper, we propose a hybrid auto-scaling approach that uses reactive and proactive solutions. We focus on the operation of a Web server, where we applied the characteristic of CPU-bound in the application. The approach proposed is a CPU usage monitor and auto-scaling management that is used to improve the system's throughput and reduce the CPU idle time. Our proposal is implemented using triggering thresholds and five forecasting models: Drift, Simple Exponential Smoothing, Holt, Holt-Winters and ARIMA. The main goal of this research is to achieve a better QoS related to the cloud computing environment. One of the obtained results shows that our method represents a throughput improvement of 12.11% by using our proposal, instead of only using a threshold-based technique.

Service providers may build elastic systems on public clouds. The public cloud may offer economies of scale, but there are some considerations to take into account. Infrastructure-as-a-Service (IaaS) providers charge their customers by the use of virtual machines (VMs), and wrong deployment decisions may lead to financial losses. This paper proposes an approach for estimating systems' performance, use of VM instances and its related costs on a public cloud. This work proposes a Stochastic Petri Net (SPN)-based formal modeling strategy to represent elastic systems deployed on a public cloud and a cost model to predict the use of VM instances. The approach enables designers to plan and tune elastic architectures based on Mean Response Time (MRT) estimation. Using our strategy it is possible to estimate the impact of each deployment configuration on the evaluated metrics. Our modeling strategy considers reactive scaling policies. The model represents a remote infrastructure for supporting external users requests. By combining different instance types, simultaneous jobs per VM instance, stepsizes and scaling thresholds, it is possible to offer different response times for each deployment configuration. One case study was performed to evaluate the approach. Our approach has proven to be feasible and it highlights the most effective scenarios to minimize MRT and reduce costs.

The infrastructure necessary to support convergent network services is increasingly more complex than the usual communication networks. A network device must be able to quickly detect any communication failure between adjacent devices for the upper-layer protocol to be able rectify this failure and prevent any interruptions in services. Convergent networks use Hello routing protocol from the upper layer to detect failures. The detection period, dependent on issues intrinsic to the protocol and the arrangement of the network configuration, can lead to high downtimes in case of unavailability of the main circuit and of switch to the backup circuit. In this context, considering a real corporate network in a production environment, there is the need to carry out experiments aimed at obtaining a scenario that can offer the lowest downtime and, consequently, a greater network availability. This work is aimed at carrying out an inferential performance assessment through a paired before-and-after comparison in order to obtain the best convergence condition for a corporate network and consequently offer greater availability. For that aim, we have proposed a Continuous-Time Markov Chain (CTMC) that represents convergent network architectures considering the Border Gateway Protocol (BGP) and Bidirectional Forwarding Detection (BFD) protocol for supporting availability evaluation. In addition, we presented a closed-form equation derived from the proposed CTMC for calculating the availability of critical warm standby components by system considering its size. The work also presents a sensitivity analysis of network availability as a result of fail over time. Our approach has proven to be feasible, and it highlights the most appropriate scenarios, supporting network architects at design time.

Distributed computing is bringing many advantages in cost, flexibility and availability. However, it increases the demand for performance and reliability. Resources such as CPU, memory, storage and network bandwidth, are very susceptible of presenting software aging issues. Therefore, proactive actions, also known as software rejuvenation must be performed to avoid these issues. The identification of the best moment to perform software rejuvenation is not a simple task, mostly because it may affect the system's availability and reliability. To overcome this problem, we propose an automatic forecasting strategy to support the system administrators to choose the best moment to perform software rejuvenation. Our strategy uses six time series techniques: Drift, Simple Exponential Smoothing, Holt, Holt-Winters, Linear Regression, and ARIMA. In our proposal, the most suitable one is chosen automatically as the best fit for a particular scenario. Three case studies were performed to evaluate the efficiency of our automatic strategy. Our proposal aims to increase the system's availability while decreasing the QoS violation probability. In one of our experiments, we can observe a reduction of 92.3% in the system's downtime. This research supports decision making activities and opens possibilities to foster the usage of forecasting strategies when dealing with software aging phenomenon.

The blockchain shared ledger emerged as an alternative to the bureaucratic banking system that may take days to confirm a payment or a transfer between clients. The blockchain concept evolved and became viable for various applications beyond the domain of financial transactions. Blockchains become a way to reach better relationships through contract validation, documents transfer, and personal and business data security. Recently, the blockchain-as-a-service has debuted on Microsoft Data Centers, and now many share an infrastructure that can change and improve their security routines. This paper evaluates the feasibility of a blockchain-as-a-service infrastructure and helps those who plan to deploy or sell blockchains. A modeling methodology based on Dynamical Reliability Block Diagrams (DRBD) is adopted to evaluate two dependability attributes: system's reliability and availability. The proposed infrastructure contains the minimum requirements to deploy the Hyperledger Cello, a platform to create and manage blockchains. The availability results pointed out a system downtime of 121 hours per year and reliability issues that must be addressed when building blockchain-as-a-service infrastructures.

The demands of performance, availability and storage of new technologies have increased significantly the energy consumption of data centers. This consumption is rising both the environmental impact and operational costs of computational systems that support those technologies. Assuming the electricity consumption and CO2emissions, the adopted utility power source is of substantial importance. This work estimates the availability, and conducts an evaluation of cost and CO2emissions of electrical infrastructures in data centers, considering different energy sources. We use a multi-layered artificial neural network, which is able to forecast consumption over the following months, based on the energy consumption history of the data center. All these features are supported by a tool, the applicability of which is demonstrated through a case study that computes the CO2emissions and operational costs of a data center using the energy mix adopted in Brazil. China. Germany and the US.

The adoption of stochastic models has been one of the central topics in various architectures. One important step to adopt it is model validation, which aims at obtaining reasonable models to represent actual behavior of services components, it has been essential to validate models against actual measurements. System-wide model simulation results can be compared with recordings from the measurement. In this paper, we accomplish the model validation to Stochastic Petri Net (SPN) models created to evaluate VoD system hosted on a private cloud system, considering MP4, MPG, Ogg and FLV formats. We proposed the performance model to represent packet transfers, and to compute performance metrics, such as throughput, packet loss, and service delivery reliability. The SPN model enables a compact representation of a large number of packets generated by video streaming. We validate the models through experimental data, using a VoD streaming service in a cloud infrastructure testbed. We demonstrate that the proposed models are accurate and can be utilized for planning the quality of service (QoS) of corporate video streaming infrastructures. A case study is presented to compare the behavior of the system. Results indicate that the model validation way adopted can be a good solution for models validation. A case study is conducted to compare the behavior of the system under distinct network scenarios. The results indicate that video streaming QoS under 3G (EVDO) networks is significantly worse than other wireless technologies, such as WiFi, 3.5G (HSPA+), and 4G (LTE).

The availability of cloud services is a critical factor for service providers because it is necessary to comply with the service level agreement (SLA) even in the event of failure. This paper proposes a strategy that uses distributed agents to choose the link with the best throughput from a faulty Data Center to an operational one. To accomplish this goal, we propose a methodology based on the analysis of network parameters (Delay and Jitter) to measure the impact of the over the time to transfer a bulk of data. Besides, we provide an algorithm that measures the RTT and the Throughput and, based on them, helping in decision-making and selecting the link with the best throughput. The agents' decision-making process is to be conducted in a distributed way using the software Apache Kafka, a consolidated open source distributed streaming platform.

Attacks that deny access to a service provider can occur anytime, anywhere, and most usually occur with little or no warning. Many small and midsize companies are not prepared to handle a significant outage. For an enterprise to face up to an attack of this type, it must possess a bandwidth higher than that of the attack, an infrastructure with redundant components, regular backups, firewalls for monitoring the threats and other proactive and reactive mechanisms. Otherwise, the service will be interrupted, increasing the chances of financial losses. Hierarchical modeling approaches are often used to evaluate the availability of such systems, thereby leveraging the representation of multiple failure and repair events in distinct parts of the system. This paper evaluates the impact of a distributed denial-of-service attack and malicious software in computer systems. We propose hierarchical models that represent the behavior of major system components and assess the effects of a DDoS and Malware attack on the system availability. We also estimate the likelihood of an attack, attacker benefits, feasibility, the pain factor and the propensity of the offense were present. They enable a direct analytical solution for large systems. The attack tree indices show the impact of simultaneous attacks on a computer system and the several threats which will maximize the system downtime. The results obtained from the attack tree analysis allow to plan and improve system's availability, maintainability, and reliability.

Mobile Cloud Computing (MCC) is a technique for increasing the performance of mobile apps and reducing their energy consumption through code and data offloading. Building an MCC infrastructure is a difficult task due to its inherent complexity and the involvement of different components. This paper proposes an approach for estimating applications' performance and data traffic volume generated by tasks offloading. This work proposes a Stochastic Petri Net (SPN)-based formal framework to represent the partitioning of applications in a method-call level. Our framework considers the available network bandwidth to send and receive tasks to the cloud. The modeling strategy represents the use and sharing of the actual available bandwidth for offloading operations. The approach enables designers to plan and tune MCC architectures based on Mean Time to Execute (MTTE) and Throughput estimation. Using our strategy it is possible to estimate the impact of the bandwidth variation on the application's MTTE and Throughput. In addition, the strategies proposed in this work may be adapted to support MCC applications in real time providing on-the-fly probabilistic performance predictions. One case study was performed to evaluate the approach. Our proposed approach has proven to be feasible and it highlights the most appropriate strategies for offloading.

Mobile Cloud Computing (MCC) helps increasing performance of intensive mobile applications by offloading heavy tasks to cloud computing infrastructures. The first step in this procedure is partitioning the application into small tasks and identifying those that are better suited for offloading. The method call partitioning strategy splits the code into a set of method calls that are offloaded to remote servers. Quite often, many applications need to make use of multiple servers for parallel processing of intensive computational operations. Predicting the behavior of such parallelizable applications is not an easy task. Deciding the number of remote servers determines the performance of the applications and the costs of the cloud usage. On one hand, users are interested in improving the performance of their applications, so they would like to use as many servers as possible, but on the other hand, they would also like to reduce their costs by using fewer cloud resources. In this paper, we propose a Stochastic Petri Net (SPN) modeling strategy to represent method call executions of mobile cloud systems. This approach enables a designer to plan and optimize MCC environments in which SPNs represent the system behavior and estimate the execution time of parallelizable applications.

The deployment of Intelligent Transportation Systems (ITS) is a challenge for industry and scientific community. Traffic characteristics change widely within a single day, therefore, ITS engineers and researchers must deal with that dynamic behavior. On the other hand, once the ITS depends on networking services, specific studies are required to consider the communication parameters along with vehicle mobility. In this paper, we propose an analytical model based on the Stochastic Petri Net (SPN) theory for evaluating Vehicular Ad-Hoc Networks (VANETs) infrastructures, considering mobility and network parameters, and the respective constraints. We employ expolynomial distributions to represent Roadside Unit (RSU) service rates. Those probability distributions allow the approximation of many analytical and empirical data. Results show that parameters such as vehicular density, message frequency, and RSU radius may affect significantly the overall system performance.

The evaluation of dependability or performance of general systems usually relies on the assistance of stochastic modeling and simulation tools. Those software packages enables the creation of models and computation of metrics quickly and accurately. This paper introduces the Mercury tool, which is an integrated software that enables creating and evaluating Reliability Block Diagrams, Stochastic Petri Nets, Continuous Time Markov Chains, and Energy Flow Models. Mercury provides a graphical user interface, a script language for command-line interface, and also an API (Application Programming Interface) that enables interaction through external applications. The evaluation of models is not restricted to the assumption of Exponential distributions, which is a common constraint in other similar tools. Mercury implements a simulation framework that allows more than 25 probability distributions, as well as a moment matching method that enables expolynomial -phase-type -distributions for models solved through numerical analysis. This paper presents the main features and methods available in Mercury to aid the dependability and performance evaluation of various systems, for both academy and industry. The accuracy and applicability of the tool is illustrated by a case study of packet loss and throughput for a Video on Demand (VoD) system.

Predicting the amount of resources available to system's users has become a task of interest to services providers even with the advent of elastic cloud computing, because the number of resources is finite despite being virtually infinite on the customer view. This paper proposes a model to evaluate node's capacity in a cloud computing environment based on the amount of available hardware resources. By combining models to availability evaluation, such as reliability block diagrams, representing the operational infrastructure mode, and stochastic Petri net for capacity-oriented availability evaluation, we can determine the real amount of resources available at a predetermined time interval. Sensitivity analysis is also used to determine the component with highest impact in the metric of interest. The models, methods, and results of this research shall aid companies to plan the deployment and configuration of their services and cloud computing infrastructure.

This short paper presents ongoing work towards the model-based evaluation of AFDX avionic networks. The required reliability of this type of network depends on timely packet delivery, which must be analyzed during design. We extend previous work by colored stochastic Petri net models of AFDX modules, which reduces several restrictions of the earlier Petri net models. Rare-event simulation for such models allows to efficiently compute small probabilities of long (potentially safetycritical) transmission delays. This has been implemented in the software tool TimeNET previously, and the simulation results validate the proposed model.

Cloud computing has the potential to reduce the cost of systems on the Internet. Elasticity mechanisms, such as auto-scaling, enable avoiding wastes, delivering only the necessary resources. Defining and implementing an efficient auto-scaling policy is a complex task that depends on the parameter setting, types of VM contracts and the expected workload. All those variables must be taken into account when establishing the tradeoffs between performance and cost to fulfill a given servicelevel agreement (SLA). We propose a stochastic model to assist in cloud planning. The model was validated for a set of significant scenarios by comparing the respective model's results with those obtained from real system measurements. This model takes as input the auto-scaling configuration parameters and the time between user requests. The proposed model is employed to compute throughput, mean response time, and cost of the cloud computing infrastructure setup. A sensitivity analysis was also conducted for identifying the parameters impact on the system performance.

The perfect relationship between deployment costs and systems availability is one of the primary goals of companies that wish to provide some computer environment or service through the Internet. The question that everyone wants to know the answer is: How much may I save and still improve the availability of my system avoiding financial losses with an SLA contract breach? This paper attempts to respond to this question by combining some techniques used at dependability evaluation field, such like hierarchical models, sensitive analysis techniques, and analysis costs. The models shown in this paper are artifacts and scenarios for a data synchronization server hosted at a private cloud computing platform based on Eucalyptus. Each scenario presented here was generated to study the impact of redundant techniques in the system availability and deployment cost. By analyzing each architecture separately and comparing them, we can indicate the one that has the best cost-benefit relationship and attends the providers, as well as the client's needs.

Users of computer systems wish to keep their personal data safe, updated, fair and accessible by other terminals, like personal computers, smart phones, portable consoles and PDAs. To perform these activities, one technology has become popular in our daily lives: data synchronization. Companies that provide this kind of service must do it with the greatest availability possible since their clients need their data to be available whenever they want to access it, and their customers in the legal field must avoid financial losses through SLA contract breaches. This paper presents hierarchical models for evaluating the availability of a data synchronization server infrastructure. The results show an availability of 98.82% for the proposed architecture, which means an annual downtime of 103 hours, this is more than 4 days of unavailability, where users cannot perform data synchronization.

Mobile Cloud Computing (MCC) combines mobile computing and cloud computing aiming to aid performance of mobile devices. The idea is simple: thin devices offload heavy methods to resource-rich servers in the clouds. We believe that in the near future MCC will adopt more advanced offloading techniques. In particular, in this paper we envision a scenario where offloading frameworks will have to deal with GPU code offloading. Amazon already offers instances with Graphics Processing Units (GPU), which can be used for this purpose. We propose and implement MCC-Adviser, a simulation tool that can predict the performance of GPUs with different number of cores using Stochastic Petri Nets. We tested MCC-Adviser in a case study with one of the expensive Amazon GPU instances. The simulations showed that it is possible to minimize costs, while satisfying user's quality of service requirements, by utilizing less powerful instances.

Elasticity is an outstanding concept of cloud computing, usually deployed through mechanisms such as auto scaling and load balancing. Cloud-based applications are able to adapt themselves dynamically to the workload behavior due to such mechanisms. The efficient instantiation of Virtual Machines (VMs) is one requirement for the elastic behavior of cloud-based applications. This study characterizes the performance of VM instantiation in a private cloud platform, considering distinct factors such as VM type, VM image size, and VM caching. We employed a full factorial design of experiments (DoE) to compute the effect and relevance of the factors as well as their interactions. Our experimental results show that the cache factor has an impact of 45.07 % on the total instantiation time, whereas the machine image (MI) has 26.45 % and the VM type only 1.05 %. The results of these experiments are also used as input parameters in a Markov chain model for sensitivity analysis. The model evaluation showed that for 6 GB and 8 GB MI, the probability of finding the MI on cache must be at least 40 % and 60 % respectively, to achieve an average instantiation time of 300 seconds. For MI with size 2 GB, such time is not exceeded even with the cache disabled. This analysis allows checking the impact of every parameter on the system response time and pointing out effective ways for improvement of performance. Such conclusions may be used as decision support for systems which often instantiate new VMs, including those using elasticity features, such as auto scaling.

High Performance Computing (HPC) aggregates computing power in order to solve large and complex problems in different knowledge areas. Nowadays, HPC users can utilize virtualized infrastructures as a low-cost alternative to deploy their applications. However, virtualization brings some challenges for HPC, specially in regard to overhead caused by hyper visors. In this work, our main goal is to analyze the performance of two hyper visors (KVM and Virtual Box) under HPC activities, considering full virtualization, and Para virtualization approaches. We used the HPC Challenge Benchmark (HPCC) to evaluate processor, RAM, inter-process communication and network communication performance. Our results show KVM in Para virtualization mode has a similar performance of a native cluster.

The increasing development and utilization of services based on cloud facilitates the development of sectors like entertainment through media streaming. Platforms such as Eucalyptus provide Infrastructure as a Service (IaaS), allowing to create and manage a wide variety of services on cloud. Redundant mechanisms are commonly adopted to achieve improvements of availability, since this issue is critical for providing a cloud service. This paper analyses the impact of adopting a redundant component for a VoD streaming service based on Eucalyptus platform, comparing these results to a non-redundant proposal. The proposed solution revealed an reduction of 51.01% in annual downtime compared to the non-redundant proposal. Sensitivity analysis was used to identify availability bottlenecks to the proposed redundant architecture.

Elasticity mechanisms allow private cloud platforms to dynamically increase or decrease resources according to policies related to workload. Auto scaling and virtual machines (VMs) instantiation are processes which affect the performance of such mechanisms. This paper evaluates the performance of the auto scaling process on a private cloud platform considering some relevant factors which are not addressed in many related research studies. Besides the experimental study, this paper presents a Markov chain model for parametric sensitivity analysis, useful to prioritize efforts in certain process parameters. The evaluation approach and results from this study can help system administrators to properly configure the auto scaling mechanism in private clouds frameworks.

Large cities face growing mobility problems, due to the major traffic jams that result from high numbers of vehicles on the roads. In response, city and national governments have invested in alternative means of urban passenger transit, such as subways, trains, as well as Bus Rapid Transit (BRT). This article aims to analyze the BRT system, by attempting to calculate the probability of reaching a destination at a specific time, thereby providing a tool that can be employed to improve the system and increase passenger confidence in it. To this end, a Continuous Time Markov Chain (CTMC) model is proposed to represent the bus stations and compute the probability metric for arrival at the destination within the specified time frame. The model allows a mathematical function to calculate the probabilities for the corresponding architecture. Two case studies were conducted in order to verify the model and illustrate its potential value in the planning of BRT systems.

The reduction of cloud computing energy consumption can be related to the application layer management with virtualization services, or a software layer. All clouds demand a power infrastructure with high availability. A data center power system is generally classified according to four redundancy levels, named tier I to IV. This paper proposes a power load distribution algorithm in depth search (PLDA-D) to optimize the power distribution of private cloud electrical infrastructures. The PLDA-D adopts the Energy Flow Model (EFM) as the basis to design power infrastructures. The EFM is a model that computes sustainability impacts and cost issues while it respects the restrictions of each component to provide energy. In addition, a case study illustrates the applicability of the proposed PLDA-D through the analysis of a private cloud running over a data center made up of tier II electrical architecture. Considerable results were achieved though the evaluation of the proposed algorithm. For instance, a reduction of 2.95% in energy consumption as well as an improvement of over 17% on the environment impact.

Cloud computing is a model in which resources such as storage, applications, and networking infrastructures can be offered as services over the internet. Cloud applications are becoming even bigger and more complex with a high availability requirement. This paper presents a modeling strategy based on a hierarchical and heterogeneous modeling for cloud infrastructure planning. This modeling strategy allows the selection of cloud infrastructures according to dependability and cost requirements. Additionally, a stochastic model generator for cloud infrastructure planning provides automatic generation of dependability and cost models for representing cloud infrastructures. A case study based on Moodle hosted on a Eucalyptus platform is adopted to demonstrate the feasibility of the proposed solution (modeling strategy and tooling).

Cloud computing is a computational paradigm in which resources such as storage, applications and networking infrastructures can be offered as services over the internet. Due to dynamic and virtualized nature of cloud environments and diversity of client requests, providing the expected service quality while avoiding over-provisioning is not a simple task. To ensure that the provisioned service is acceptable, providers must exploit techniques and mechanisms that guarantee a minimum level of service quality. This performance evaluation of cloud infrastructures has been receiving considerable attention by providers as a prominent activity for improving service quality. This paper presents a modeling strategy for cloud infrastructure planning with different software and hardware configurations, according to performance and cost requirements. A case study based on Moodle hosted on Eucalyptus platform is adopted to demonstrate the feasibility of the proposed modeling strategy.

Cloud computing is a new paradigm that provides services through the Internet. Such paradigm has the influence of the previous available technologies (e.g., cluster, peer-to-peer and grid computing) and has been adopted to reduce costs, to provide flexibility and to make management easier. Companies like Google, Amazon, Microsoft, IBM, HP, Yahoo, Oracle, and EMC have conducted significant investments on cloud infrastructure to provide services with high availability levels. The advantages of cloud computing allowed the construction of digital libraries that represent collections of information. This system demands high reliability and studies regarding analysis of availability are important due to the relevance of conservation and dissemination of the scientific and literature information. This paper proposes an approach to model and evaluate the availability of a digital library. A case study is conducted to show the applicability of the proposed approach. The obtained results are useful for the design of this system since missing data can lead to various errors and incalculable losses.

Cloud computing environments provide powerful storage and processing capabilities, as well as other computational resources. Due to the business potential of the pay-peruse model, as well as the advantages of easy scalability, up-to-date Video on Demand (VoD) streaming services rely on cloud infrastructures to offer a wide variety of multimedia content. This paper proposes the application of availability models to a cloud environment designed for a video streaming service. Sensitivity analysis is proposed to identify the availability bottlenecks.

Use of server virtualization for providing applications produces overloads that degrade the performance of provided systems. The use of container-based virtualization enabled a narrowing of this overload. On this work, we go a step forward and demonstrate how a broad tuning combination of several performance factors concerning to web cache server - the I/O bound application analysed - to file system and to operating system, led to a higher performance of proposed cluster, when it is executed on a container-based operating system virtualization environment. Availability and performance similarity of web cache service, under non-virtualized and virtualized systems, were evaluated when submitted to proposed web workload. Results reveal that web cache service provided under virtual environment, without unresponsiveness fails due to overload, i. e., with high availability, presents a 6% higher hit ratio and a 21.4% lower response time than those observed on non-virtualized environments.

The use of mobile cloud computing infrastructure can be used to improve the delivery of health services. This paper proposes a high-level model that characterizes the behavior of an mHealth system. The main objective of this work is to identify the probability of a message being delivered at a time t, whereas several components involved may fail. Various scenarios were analysed considering different possibilities for communications, battery discharge rates and timeouts. The results show that, in the best case, a message may be delivered in 1.3 minutes with a 0.9999 probability, but in the worst case may take up to 8.8 minutes.

Cloud computing environments are an emerging technology that aims at proving several cloud services over the Internet, making it possible to provide a huge variety of applications with different purposes to users. Multimedia services are examples of these new services that use cloud computing, like video streaming, where the user can access their videos from cloud environments. For cloud service systems, assurance of high-availability of cloud services is a challenging and critical issue. In this paper, we evaluated the video streaming service's availability for video on demand cloud system. Hierarchical modeling strategy, availability models combining reliability blocks diagrams (RBD) and continuous time Markov chain (CTMC) were used, indicating a system availability of 1.92 nines. The critical component of the system was indicated through sensibility analysis. We realized a model validation technique to demonstrate that the models represent the behavior of the real system.

Managing a cloud computing provider is a difficult task, which involves the control and maintenance of several components, such as computers, network infrastructures and software components. In these environments, availability, security and low costs are important requirements to achieve high quality of service. Therefore, the evaluation of these systems is important to find a configuration that meets the constraints of users and provider. A widely adopted strategy to evaluate cloud computing systems consists by the utilization of stochastic models (e.g., stochastic Petri nets - SPN) to assess the concern metrics. In this context, Eucalyptus is an open source private cloud software for building private and hybrid clouds. This work presents dependability models for evaluation on Eucalyptus clouds. These models focus on the user point of view metrics (e.g., number of running virtual machines) to assess the dependability metrics. In order to demonstrate the feasibility of the proposed models, we evaluate a real world environment and validate the presented models by using Eucabomber tool version 2.0.

Over the last years, many companies have been moving their computational services to cloud computing providers. In these environments, high availability is a key metric since system outages can lead penalties defined in service level agreements (SLAs). Eucalyptus is a prominent tool for providing virtual machines (VMs) as a cloud service. Considering the user perspective, the service is available if the requested VMs are operational no matter which part of the infrastructure is failed or not. Therefore, providers need to evaluate the dependability level of their environment, considering the user point of view. An important technique to perform dependabiliy evaluation in general systems corresponds to fault injection approach, which observes the systems behavior during the occurrence of fault events. This paper presents an extension of previous work (i.e., Eucabomber) that provides a fault injection tool for dependability evaluation in Eucalyptus cloud computing platform. The current version extends the previous one by providing new user-oriented metrics considering the VM life-cycle operations and also takes into account failure dependencies between Eucalyptus components. Additionally, a case study comparing dependability model results and measured metrics is provided to demonstrate the feasibility of the proposed environment.

Distributed storage systems (DSS) are becoming even more important and complex due to the emergence of paradigms such as cloud computing, and the challenges for dealing with big data. DSS provide a set of services such as publishing, file sharing and high performance utility storage. In addition, there is a growing demand to support these services with high scalability, availability and low cost. Sheepdog is a distributed storage system which aims at providing storage volumes with high availability, using clusters of common machines. This type of storage system may prove useful in the VMS (Video Monitoring Service) market and as remote storage for cloud-based applications. This paper analyzes the scalability of Sheepdog clusters, by measuring the performance with different number of nodes and different replication degrees. The system analyzed here showed better performance for read operations in small cluster sizes than in larger ones. On the other hand, proper scalability was found for the write throughput, since no significant performance degradation was noticed when the number of cluster nodes increased.

Disasters can strike any time, anywhere, and most usually occur with little or no warning. Many small and mid-size businesses (SMBs) are not adequately prepared to handle a major outage. A potential solution is the the use of cloud computing to enable a disaster recovery solution for SMBs, alleviating the burden of allocating resources which would be used only in a catastrophic situation. This paper describes an experiment-based approach for studying the dependability of a disaster recovery solution supported by a private cloud. We employ fault injection to verify the benefits and drawbacks of such a solution during the design and early deployment phases. Dependability measures obtained in the testbed can be used to adjust specific components of the solution, to cross-check analytical and simulation models, as well as to give foundations for the definition of service level agreements with customers.

Measurement is an important activity whenever design embedded systems, as it provides means for platform characterization and selection as well as system validation and specification. The sample frequency is a key aspect for obtaining accurate and precise measures. Obtaining highly accurate and precise measures may require measurement systems able to gauge the system under measurement at very high sample rates. The sample rate and the sensing device of a measurement system are features that decisively affect its prices. Variability and noise are intrinsic phenomena to be handled, thus the adoption of robust and fast statistic methods are indispensable for accuracy, precision and may favour the cost reduction of measurement systems. This paper adopts a low cost measurement system for gauging execution time and energy consumption of embedded software applications. This platform allows designers to obtain accurate and precise estimates of software piece of codes as well as it is an attractive means for supporting embedded software code tuning.

Cloud computing is a model that resources such as storage, applications and networking infrastructures can be offered as services over the internet. Cloud applications are becoming ever larger and more complex with high availability requirement. This paper presents a modeling strategy based on a hierarchical and heterogeneous modeling for planning of cloud infrastructures. The modeling strategy allows the selection of cloud infrastructures according to dependability and cost requirements. A case study based on virtual learning environments hosted on Eucalyptus platform is adopted to demonstrate the feasibility of the proposed modeling strategy.

The need for high reliability, availability and performance has significantly increased in modern applications, that handle rapidly growing demands while providing uninterruptible services. Cloud computing systems fundamentally provide access to large pools of data and computational resources. Eucalyptus is a software framework largely used to implement private clouds and hybrid-style Infrastructure as a Service. It implements the Amazon Web Service (AWS) API, allowing interoperability with other AWS-based services. This article investigates the software aging effects in the Eucalyptus framework, considering workloads composed of intensive requests for remote storage attachment and virtual machine instantiations. We found problems that may be harmful to system dependability and performance, specifically regarding to RAM memory and swap space exhaustion, besides highly excessive CPU utilization by the virtual machines. We also present an approach that applies time series analysis to schedule rejuvenation, so as to reduce the downtime by predicting the proper moment to perform the rejuvenation. We experimentally evaluate our approach using an Eucalyptus test bed. The results show that our approach achieves higher availability, when compared to a threshold-triggered rejuvenation method based on continuous monitoring of resources utilization.

In this work, the problem of design space exploration of soft real-time embedded systems is formulated as a stochastic simulation optimization problem. A novel multiobjective genetic algorithm is proposed to address this problem. In the proposed algorithm, design metrics such as price and size are optimized while deadline violations are minimized. Experimental results show the advantages of our approach over previous work that consider no deadline violation is tolerable.

With the increasing adoption of cloud computing environments, studies about high availability in those systems became more and more significant. Software rejuvenation is an important mechanism to improve system availability. This paper presents a comprehensive availability model to evaluate the utilization of the live migration mechanism to enable VMM rejuvenation with minimum service interruption. Live migrations are performed observing a time-based trigger. We evaluate five different scenarios, with distinct time intervals for triggering the rejuvenation. The results show that the live migration can significantly reduce the system downtime.

Hundreds of natural disasters occur in many parts of the world every year, causing billions of dollars in damages. This fact contrasts with the high availability requirement of cloud computing systems, and, to protect such systems from unforeseen catastrophe, a recovery plan requires the utilization of different data centers located far enough apart. However, the time to migrate a VM from a data center to another increases due to distance. This work presents dependability models for evaluating distributed cloud computing systems deployed into multiple data centers considering disaster occurrence. Additionally, we present a case study which evaluates several scenarios with different VM migration times and distances between data centers.

This paper proposes an investigative approach to indicators of software aging in applications developed for Android, a Linux-based operating system primarily designed for touch screen mobile devices such as smartphones and tablets. Software aging is a degrading factor in systems, leading to poor performance, failures, and may ultimately result in system downtime. The methodology proposed identifies memory leaking in Android applications. In order to test the approach, case studies were performed with the Monkey tool as workload generator, employing Linux utilities to monitor the environment. Communication between the computer and the smartphone was handled with the ADB tool, a toolkit included in the Android SDK package. Experimental results confirmed both the effectiveness of the procedure and the existence of software aging phenomenon in the Foursquare application running on the Android OS platform.

Cloud computing has expanded in recent years due to many effects, such as accessibility, reliability, and collaboration. To provide those functionalities high availability is in demand, which implies a higher electric energy consumption by the computers that support the cloud infrastructure. Studies that pay attention to this electric energy consumption are important due to its impact on sustainability and operational costs. This paper proposes a power load distribution algorithm (PLDA) to optimize electrical flows of power infrastructures. The PLDA adopts the Energy Flow Model (EFM) as its basis. The EFM is a model that computes sustainability impacts and cost issues, while it respects the energy providing restrictions of each component. In addition, a case study illustrates the applicability of the proposed PLDA through the analysis of six private cloud power architectures. Considerable results were observed, including a reduction on energy consumption of 10.7%, and an improvement (reduction) on the environmental impact of over 140% was obtained.

Software aging refers to performance degradation faced on long-running execution software. The software infrastructure supporting Cloud Computing systems may suffer from aging effects. Such a phenomenon was already reported for virtual infrastructure managers and virtual machine monitors. In this sense, software rejuvenation is a technique of proactive fault tolerance to deal with effects of software aging. This paper proposes a comprehensive availability model for two different types of rejuvenation scheduling based on live migration mechanism, one with a test before migration, and other without. We evaluate five different scenarios, with distinct time intervals for triggering the rejuvenation. The main goal is to explain the benefits gained from the use of this type of rejuvenation technique, as well as understand the differences between the two approaches analyzed. The results show that the utilization of a schedule with a checking mechanism before rejuvenation can bring a significant improvement on system availability.

Mobile cloud computing refers to abstraction of computational power to outside of the mobile device, processed in the "clouds". This paper provides an availability and energy consumption study using hierarchical heterogeneous modeling. Different scenarios were analyzed considering wireless communication technologies, such as 3G and WiFi. The results shows that the 3G protocol causes a lower availability and is the greater responsible for the discharge of battery, but when is combined with WiFi protocol providing communication redundancy, we have the best results of availability and better results of uninterrupted operation time of battery.

Emergency systems serve people in utmost circumstances, hence they should be highly dependable. Availability and performance are key aspects in emergency systems. Architectural models, service policies, and redundancies are attributes for defining and evaluating the overall operation of such systems. In this work, we performed the perform ability evaluation of an important emergency system located in a large city in Brazil. The results showed that the system's downtime is considered high, so, we proposed a new architecture to improve the system's perform ability. The results obtained in this paper can be used to provide support for decisions on interventions in the emergency system to improve its perform ability. It is expected that the models presented in this paper will be useful in a variety of emergency systems.

Over the last years, the evaluation of cloud computing infrastructures has received considerable attention as a prominent activity for improving service quality as well as planning modifications in an existing infrastructure. This paper presents Stochastic Model Generator for Cloud Infrastructure Planning (SMG4CIP), a automatic generation system of stochastic models for cloud infrastructure planning. This system indicates feasible cloud infrastructures according to dependability and cost requirements. The proposed system adopts a technique based on GRASP metaheuristic in order to recommend optimized cloud infrastructures. Furthermore, SMG4CIP adopts stochastic models, such as Petri Net and Reliability Block Diagram, to evaluate dependability and cost metrics. A case study based on Electronic Funds Transfer (EFT) system is adopted to demonstrate the feasibility of the proposed system.

Data centers are constantly growing in order to meet the demands of new technologies such as cloud computing and e-commerce. In such paradigms, periods of downtime can lead to financial losses of millions of dollars and permanently damage a company's reputation. Several factors affect the availability of IT systems in data centers, among them temperature variations within the data center room. This paper proposes an approach for evaluating the effect of such temperature changes on data center IT systems. In this approach, the evaluation is accomplished with stochastic Petri net (SPN) models which simulate the impact of temperature on the availability of variously specified IT and cooling architectures. A case study is included to demonstrate the applicability of the proposed models.

Cloud computing is a computational paradigm with increasing adoption because it offers resources as services in a dynamically scalable way through the Internet. The constant concern in providing cloud computing services in a reliable and uninterrupted manner inspires availability and reliability studies. A feasible method of performing such studies is through automated fault injection, enabling to observe the behavior of the cloud architecture under many conditions. This paper presents a fault injection tool, named EucaBomber, for reliability and availability studies in the Eucalyptus cloud computing platform. EucaBomber allows to define the probability distribution associated to the time between generated events. The efficiency of EucaBomber is verified through testbed scenarios where faults and repairs are injected in a private Eucalyptus cloud. The experimental results are cross-checked with results estimated from a Reliability Block Diagram, using the same input parameters of the experimental testbed. The test scenarios also illustrate how the tool may assist cloud systems administrators and planners to evaluate the system's availability and maintenance policies.

Customer requirements and local legislation are forcing companies to provide green products and services. Such driving forces have resulted in the study of green supply chains (SCs), which are SCs that take into account factors that may cause positive or negative environmental impacts. However, some environmental indicators are often at odds with business goals, leading to a conflict, which, in order to optimize several goals simultaneously, is usually resolved by utilizing multicriteria decision analysis. Other modeling techniques describe systems by considering temporal aspects and causal relations. Such is the case of Petri nets, which were previously widely used to analyze performance in the context of SCs. This paper presents a method that combines the strengths of both techniques to assess the environmental and business indicators of such systems. This method adopts the exergy and the global warming potential key environmental indicators. A case study is conducted to demonstrate the application of the proposed modeling technique.

One of the key characteristics of virtual networks (VN) is their intrinsic dynamical aspects. VNs allow network operators to have on-demand negotiation of a variety of services. Risks are inherent to virtualized infrastructures since the underlying physical network components are failure-prone. Therefore evaluation and analysis of risks in VNs, from assessment of dependability attributes, are of paramount importance for accurate management and control tasks. This paper proposes and evaluates a method to assess dependability in virtualized network environments. Through the use of Reliability Block Diagram (RBD) and Stochastic Petri Nets (SPN), we provide a first look at risks in network virtualization by means of dependability modeling and analysis. Our methodology allows the networking research community to develop new heuristics when dealing with resource allocation in VNs, while taking into account dependability attributes.

Accelerating the time to market for new embedded real-time systems is an imperative strategy nowadays. However, releasing poor-quality designs to the market can be dangerous, since a missed deadline in hard real-time systems can be catastrophic. Thus, early detection of potential problems in these systems is mandatory, since it may reduce the risks of fault propagations from early specification to the final code. This paper presents Calau, an environment for modeling and analyzing embedded real-time systems. Calau supports the mapping process of SysML State Machine diagram into a Time Petri Net with Energy consumption in order to analyze timing and energy consumption requirements in the early phases of embedded real-time systems development life cycle. It allows designers who do not have expertise in formal model, to design and analyze embedded system specifications quantitatively and qualitatively. We show the applicability of our tool by detailing an example.

High availability in cloud computing services is essential for maintaining customer confidence and avoiding revenue losses due to SLA violation penalties. Since the software and hardware components of cloud infrastructures may have limited reliability, fault tolerance mechanisms are a means of achieving the necessary dependability requirements. This paper investigates the benefits of a warm-standy replication mechanism in a Eucalyptus cloud computing environment. A hierarchical heterogeneous modeling approach is used to represent a redundant architecture and compare its availability to that of a non-redundant architecture. Both hardware and software failures are considered in the proposed analytical models. The results show an enhanced dependability for the proposed redundant system, as well as a decrease in the annual downtime. The results also demonstrate that the simple replacement of hardware by more reliable machines would not produce improvements in system availability to the same extent as would the fault tolerant approach.

Nowadays society has increasingly relied on convergent networks as an essential means for individuals, businesses, and governments. Strategies, methods, models and techniques for preventing and handling hardware or software failure as well as avoiding performance degradation is, thus, fundamental for prevailing in business. Issues such as operational costs, revenues and the respective relationship to key performance metrics are central for defining the required system infrastructure. Our work aims to provide system performance and dependability models for supporting optimization of infrastructure design aiming at business oriented metrics. In addition, a methodology is also adopted to support both the modeling and the evaluation process.

The need for reliability, availability and performance has increased in modern applications, which need to handle rapidly growing demands while providing uninterrupted service. Cloud computing systems fundamentally provide access to large pools of data and computational resources. Eucalyptus is a software framework used to implement private clouds and hybrid-style Infrastructure as a Service. It implements the API Amazon Web Service (AWS), allowing interoperability with other AWS-based services. Elastic block storage is a technology which provides flexible allocation of remote storage volumes to the virtual machines running in a cloud computing environment. This work investigates the software aging effects on the Eucalyptus framework, considering workloads composed of intensive requests for attaching remote storage volumes to virtual machines. The results evidenced problems that may be harmful to system dependability and its performance due to RAM memory exhaustion and subsequent use of swap memory, besides high CPU utilization by the virtual machines and subsequent increase in the response time of applications running on the VMs.

The cloud computing paradigm provides virtualized computing resources as on-demand services. This paradigm allows companies to focus on their business issues rather than conceiving and managing complex infrastructures. As a consequence, performance evaluation of cloud computing infrastructures has been receiving considerable attention by cloud service providers as a prominent activity for improving service quality as well as planning modifications in the current infrastructure to support future demands. This paper focus on the capacity planning of processing and storage resources related to an Electronic Funds Transfer service hosted on the Amazon EC2 cloud infrastructure.

The scientific community has devoted attention to virtualization in hard real-time systems, motivating the development of many hierarchical scheduling techniques for dealing with stringent timing constraints for such systems. Most techniques are based on runtime scheduling methods and they provide important results concerning schedulability analysis. However, there are situations in which runtime methods may fail in finding a feasible schedule, and intertask relations as well as overheads are neglected by several approaches. This paper proposes a hierarchical pre-runtime scheduling for hard real-time systems taking into account intertask relations and overheads. A formal model based on time Petri net (TPN) is adopted to provide a basis for precise schedule generation.

Server virtualization is a technology used in many enterprise systems to reduce operation and acquisition costs, and increase the availability of their critical services. Virtualized systems may be even more complex than traditional nonvirtualized systems; thus, the quantitative assessment of system availability is even more difficult. In this paper, we propose a sensitivity analysis approach to find the parameters that deserve more attention for improving the availability of systems. Our analysis is based on Markov reward models, and suggests that host failure rate is the most important parameter when the measure of interest is the system mean time to failure. For capacity oriented availability, the failure rate of applications was found to be another major concern. The results of both analyses were cross-validated by varying each parameter in isolation, and checking the corresponding change in the measure of interest. A cost-based optimization method helps to highlight the parameter that should have higher priority in system enhancement.

The design, deployment and execution of business process models and their associated security models is expensive and time consuming. This is because these activities usually involve multiple stakeholders that include business domain experts, security experts, web service developers and IT operations teams, and there is no streamlined development environment to allow these stakeholders to work collaboratively on a business process. We have developed a cloud-based model-driven development and execution environment called SSC4Cloud to provide a shared business process modeling workspace and a business process execution environment. More specifically, with the shared modeling workspace, business process models can be developed, refined and shared. Within the shared execution environment, a business process model is translated into a WS-BPEL based executable model, which is then assigned for execution in a virtual machine container from a shared machine cluster. The common model execution environment supports both business process execution and enforcement of the security requirements attached to the business process models.

Research advances in network traffic forecasting are continually emerging in the literature. Although the progress in this field, the quality control of traffic forecasts is commonly neglected in previous works. In this paper we analyze the performance of CuSum, Shewhart, and EWMA control charts applied to the residuals of forecasting models fitted to selected traffic samples obtained from real IP networks. Based on numerical and graphical analyses, we verify that CuSum and Shewhart control charts perform worse than EWMA, since they report false alarms for several forecasting processes considered under control. Differently, the EWMA control chart shows efficacy to monitor the different types of studied forecast residuals. Finally, we present a sensitive analysis of the EWMA model with respect to the lambda parameter in order to evaluate its flexibility on different traffic scenarios.

Cloud Computing, Business Process Modeling (BPM) and Service-oriented architectures (SOA) are playing a relevant role in the evolution of Information Technology (IT). A considerable number of system developers are using Cloud technologies to deploy and make available systems over the Internet. Business Process Management standards are being widely used to model business requirements. In addition, SOA-based systems are considered an interesting approach to execute high-level business process specifications. Based on the fact that business processes are executed, usually, using services available in network environments, security requirements should be considered, especially when dealing with sensitive data (e.g., credit card information or personal data). Despite the increasing need for specifying security mechanisms in web service compositions in the Cloud, this topic remains a challenge for many reasons, including the known difficulty of expressing security requirements at a business level and the enforcement of such requirements at an execution level in a cloud environment. This work presents an environment to collaboratively model business processes considering security requirements and to automatically deploy them in the Cloud with security requirements enforcement. The business process is realized through the utilization of web service composition. This environment consists of a set of tools to support the business process modeling and secure service composition execution in the Cloud. Security-related information can be shared among different users in the Cloud and used to enable the activation and configuration of security mechanisms. The proposed approach is showcased in a Virtual Travel Agency scenario to show its feasibility.

Demands on software reliability and availability have increased due to the nature of present day applications. Cloud computing systems fundamentally provide access to large pools of data and computational resources through a variety of interfaces similarly to existing grid and HPC resource management and programming systems. This work investigates the software aging effects on the Eucalyptus cloud computing infrastructure considering workloads composed of provisioning different types of virtual machines.

Business service through the Internet, particularly the cloud computing paradigm, have demanded more computational resources from data centers to provide high-availability services. As consequence, issues such as operational costs and sustainability increase, mainly, due to the redundant infrastructures needed to support the data center operation. However, data center designers have slightly information on the sustainability impact, operational cost and availability of data center architectures. In this context, models are important to support data center designers to estimate the environmental impact, dependability as well as the cost associated to the infrastructure before implementing it. This paper presents a set of formal models for estimating sustainability impact and dependability metrics, supported by an integrated environment, namely, ASTRO.

The number and size of data centers have greatly increased in recent years, mainly, due to the explosive growth of Internet services, such as those based on cloud computing paradigm. Therefore, service providers have intensified efforts on improving the quality of services while minimizing the costs of data centers, for instance, those associated to out-of-services periods. In other words, robust data center infrastructures are necessary, such that dependability requirements are satisfied while simultaneously incurring costs as minimal as possible. This paper proposes an approach that combines reliability importance indexes with TCA (Total Cost of Acquisition) to evaluate data center power infrastructures. The proposed approach aims at indicating the benefits and drawbacks of different architectures for power infrastructures to guide the decision maker in evaluating the most appropriate design.

Traffic engineering in computer and communication networks has become a very complex process. An important part of this activity is to make predictions about network traffic. Therefore, research advances in network traffic forecasting are continually emerging in the literature. Although its importance, the quality control of network traffic forecasts is commonly neglected. In this paper we analyze the performance of CUSUM, Shewhart, and EWMA control charts applied to the residuals of an ARMA(2, 2) forecasting model processing real WAN and LAN traffic samples. Numerical results and graphical analysis indicate that CUSUM performs worst for all traffic samples. Shewhart chart performs better than CUSUM, but worse than EWMA since it reports false alarms for processes under control. Differently than CUSUM and Shewhart, EWMA (λ=0.1) control chart shows the best performance for all traffic patterns, showing no false alarms.

Software aging is a phenomenon defined as the continuing degradation of software systems during runtime, being particularly noticeable in long-running applications. Aging-related failures are very difficult to observe, because the accumulation of aging effects usually requires a long-term execution. Thus, collecting a statistically significant sample of times to aging-related failures so as to estimate the system's lifetime distribution is a very hard task. This is an important problem that prevents many experimental and analytical studies, mainly those focused on modeling of software aging aspects, of using representative parameter values. In this paper we propose and evaluate the use of quantitative accelerated life tests (QALT) to reduce the time to obtain the lifetime distribution of systems that fail due to software aging. Since QALT was developed for hardware failures, in this paper, we adapt it to software aging experiments. We test the proposed approach experimentally, estimating the lifetime distribution of a real web server system. The accuracy of the estimated distribution is evaluated by comparing its reliability estimates with a sample of failure times observed from the real system under test. The mean time to failure calculated from the real sample falls inside the 90% confidence interval constructed from the estimated lifetime distribution, demonstrating the high accuracy of the estimated model. The proposed approach reduces the time required to obtain the failure times by a factor of seven, for the real system investigated.

The growth of engineering disciplines, such as maintainability and reliability, have been motivated by several factors, which include the ever growing of system complexity as well as the high cost associated to failures repairing activities and productivity reduction. This work adopts a methodology, which includes a hybrid modeling technique that considers the advantages of both stochastic Petri nets (SPN) and reliability block diagrams (RBD) to evaluate maintenance policies with different service level agreement (SLA). Dependability, cost and sustainability are also prominent features considered in the proposed methodology.

In availability predictive analysis not every model input parameter is known exactly when the model is developed. The sources of uncertainties include lack of data and lack of knowledge about physical systems. In this paper, we present a reliable mechanism to help improve evaluation robustness when significant uncertainties exist. The mechanism incorporates variabilities and uncertainties based on imprecise distribution parameters in the availability evaluations are intervals instead of precise real numbers. The analysis takes into account the effects of variabilities in MTTR and MTTF when calculating interval availabilities. Interval studies of availability models should be used to show the influence of input parameters uncertainties on resultant availability evaluation. Interval methods and interval arithmetics have been applied in Scilab toolbox Int4Sci framework for steady state solution of availability models.

This paper describes the development and use of an interactive graphical user interface (GUI) for a synthetic trace generator named WGCap (Workload Generator for Capacity Advisor). The WGCap GUI generates traces to be imported into the Capacity Advisor, a capacity planning tool inside the HP Virtual Server Environment - VSE, for simulating the consumption of resources like CPU demand, memory size, disk data rate and network data rate. The workload generation acts in a flexible way, allowing the generation of workloads based on actual traces or on a significant set of probability distributions and statistical summaries.

In this paper, we investigate the performance and dependability modeling of voice and data services in computer networks. We use Stochastic Petri Net as an enabling modeling approach for analytical evaluation of complex scenarios. Our goal is to analyze the performability of a network infrastructure, by considering the influence of queuing policies and network topologies on the quality of voice and data services. We apply our proposed modeling approach in a case study to evaluate the dependability and performability of an enterprise network.

One of the challenges the IT industry faces involves maintaining the quality of the developed product within a consistent planning. Product delays are in general consequences of badly conceived planning and lack of progress control. The interest in software testing activities has been growing over the last few years. Software companies currently seek low cost testing procedures and at the same, with great capacity to handle mistakes. In this article, concepts related to testing processes are mapped in Stochastic Petri Nets(SPN) which provides a formal representation for measures used in performance analysis.

The advent of cloud computing has demanded more computational resources from data centers in order to provide high-availability services required in this new paradigm. To support data center dependability evaluation, this paper presents a tool, namely, ASTRO, which adopts a hybrid modeling approach, that includes Reliability Block Diagrams (RBD), Stochastic Petri Nets (SPN) and Data Center High-Level models. In addition, this tool also provides a prominent functionality in which an evaluation result can be reused in other models. ASTRO Tool also permits a hierarchical modeling, supporting the piecewise approach proposed by the methodology in which the tool is inserted.

The Service-Oriented Architecture (SOA) has become a unifying technical architecture that may be embodied through Web Service technologies. Predicting the variable behavior of SOA systems can mean a way to improve the quality of the business transactions. This paper proposes a simulation modeling approach based on stochastic Petri nets to estimate the performance of SOA-applications. Using the proposed model is possible to predict resource consumption and service levels degradation in scenarios with different compositions and workloads, even before developing the application. A case study was conducted in order to validate our approach, comparing its accuracy with the results from an analytical model existent in the literature.

This paper proposes a methodology for modeling and evaluating supply chains based ongeneralized stochastic Petri net(GSPN) components. The proposed modeling process is based on a bottom-up approach, which assures desirable model properties that start from a set of predefined modules for typical supply chain entities. A compositional algebra that formally defines GSPN compositions is presented. The proposed methodology is supported by an implementation of the methodology in thestochastic logistics optimizer tool(SLOT). An industrial case study has been conducted to show the benefits of the approach.

In this paper we use the Activity diagram of the System Modeling Language (SysML) in combination with the new UML profile for Modeling and Analysis of Real-Time and Embedded systems (MARTE) in order to validate functional, timing and low power requirements in early phases of the embedded system development life-cycle. However, SysML lacks a formal semantics and hence it is not possible to apply, directly, mathematical techniques on SysML models for system validation. Thus, a novel approach for automatic translation of SysML Activity diagram into Time Petri Net with Energy constraints (ETPN) is proposed. In order to depict the practical usability of the proposed method, a case study is presented, namely, pulse-oximeter. Besides, the estimates obtained (execution time and energy consumption) from the model are 95% close to the respective measures obtained from the real hardware platform.

The performance evaluation of Electronic Funds Transfer (EFT) Systems has an enormous importance for Electronic Transactions providers, since the computing resources must be efficiently used in order to attain requirements defined in Service Level Agreements (SLA). Among such requirements, we may stress agreements on availability, reliability, scalability and security. In the EFT system, faults can cause severe degradation on system performance. The modeling of these systems, ignoring the dependability effects on performance can lead to incomplete or inaccurate results. This paper presents an expolinomial stochastic model for evaluating the performance of the EFT system processing and storage infrastructure considering load variation range. This work also presents a model for evaluating the dependability of the EFT system infrastructures and combines both models (dependability and performance) for evaluating the impact of dependability issues on the system performance. The performability analysis employs a hierarchical decomposition method aiming at reducing the computational effort and avoiding stiffness problems. Finally, a case study is presented and the respective results are depicted, stressing important aspects of dependability and performance for EFT system planning.

The electronic funds transfer (EFT) systems promote both financial benefits for the institutions that use them and potential conveniences for their customers. EFT Systems debiting and crediting individual bank accounts from points of sale (POS) terminals in different companies and providing cash on demand 24 hours a day through automatic tellers. In the EFT system, eventual workload peaks and fluctuations can cause degradation on system performance. Hence, the performance evaluation of EFT systems has an enormous importance for electronic transactions providers, since the computing resources must be efficiently used in order to ensure high availability, reliability, scalability and security. This paper presents an stochastic model for performance evaluation and planning EFT systems processing and storage infrastructure considering load variation range. A case study is presented in order to describe the adequacy of presented approach and results.

This paper presents a performance and energy consumption modeling technique for embedded systems. The proposed method adopts a formal model based on Coloured Petri Nets for modeling the functional behavior of processors and memory architectures at a high-level of abstraction. The applicability of the proposed method is illustrated by evaluating a set of applications and a general-purpose microcontroller. Experimental results demonstrate an average accuracy of 96\% in comparison with the respective measures acquired from the real hardware platform. Moreover, the high-level behavioral representation of platforms allows the rapid analysis of performance and energy consumption of complex systems.

Capacity planning tools aim at monitoring and evaluating computational infrastructures resources in terms of workloads. The capacity advisor is a capacity planning tool inside the HP virtual server environment - VSE. This paper describes the development and use of a synthetic workload generator named WGCap (workload generator for capacity advisor), a flexible synthetic workload generator that generates traces to be imported into the capacity advisor for simulating the consumption of resources like CPU demand, memory size, disk data rate and network data rate. The workload generation is reliable and flexible, allowing the generation of workloads based on actual traces or on a significant set of probability distributions and statistical summaries.

The performance evaluation of EFT systems has paramount importance for electronic transactions providers, since the computing resources must be efficiently used in order to attain requirements defined in service level agreements (SLA). Among such requirements, we may stress agreements on availability, reliability, scalability and security. In the EFT system, faults can cause severe degradation on system performance, and as a result modeling can lead to incomplete or inaccurate results ignoring the dependability effects on performance. This paper presents an expolinomial stochastic model for evaluating the performance of the EFT system processing and storage infrastructure considering a load variation range. This work also presents a model for evaluating the dependability of the EFT system infrastructures and combines both models (dependability and performance) for evaluating the impact of dependability issues on the system performance. The performability analysis employs a hierarchical decomposition method aiming at reducing the computational effort and avoiding stiffness problems.

Dynamic voltage scaling (DVS) has been largely adopted as an effective technology for reducing energy consumption in embedded systems. Since the usage of DVS may affect the timing constraints of a hard real-time system, over the last decade, several pre-runtime as well as runtime scheduling approaches have been developed to tackle such an issue. Nevertheless, both have drawbacks that can be mitigated using a joint approach. This paper proposes a hybrid DVS scheduling approach for energy-constrained hard real-time systems, taking into account overheads, precedence and exclusion relations. The proposed method adopts a formal model based on time Petri nets in order to provide feasible schedules that satisfy timing and energy constraints.

The service-oriented architecture (SOA) has become an unifying technical architecture that can be embodied with Web service technologies, in which the Web service is thought as a fundamental building block. This paper proposes a simulation modeling approach based on stochastic Petri nets to estimate the performance of SOA applications. Using the proposed model it is possible to predict resource consumption and service levels degradation in scenarios with different compositions and workloads. A case study was conducted to validate the approach and to compare the results against an existing analytical modeling approach.

In this paper, we introduce the ezRealtime project, which relies on the Time Petri Net (TPN) formalism and defines a Domain-Specific Modeling (DSM) tool to provide an easy- to-use environment for specifying Embedded Hard Real-Time (EHRT) systems and for synthesizing timely and predictable scheduled C code. Therefore, this paper presents a generative programming method in order to boost code quality and improve substantially developer productivity by making use of automated software synthesis. The ezRealtime tool reads and automatically translates the system's specification to a time Petri net model through composition of building blocks with the purpose of providing a complete model of all tasks in the system. Hence, this model is used to find a feasible schedule by applying a depth-first search algorithm. Finally, the scheduled code is generated by traversing the feasible schedule, and replacing transition's instances by the respective code segments. We also present the application of the proposed method in an expressive case study.

The capacity management of a Java Enterprise Edition application involves the tuning of numerous parameters at different layers that include the database management system, network, operating system, Java virtual machine, JEE application server and application code level itself. At the application server level, an important element to be considered is the instance pooling mechanism, which manages the creation of instances required to process client requests. This paper proposes Petri net models to represent the instance pooling mechanism embedded in the JBoss application server and demonstrates how those models can be used for performance evaluation.

Over the last years, the issue of reducing energy consumption in hard real-time embedded systems has received considerable attention from the scientific community, since responsiveness and low energy consumption are often conflicting requirements. In this context, several techniques have been proposed, mostly, adopting DVS (dynamic voltage scaling) for reducing energy consumption in conjunction with runtime scheduling policies for dealing with stringent timing constraints. Nevertheless, those approaches generally disregard inter-task relations and runtime overheads (e.g. voltage switching) as well as adopt specialized operating systems. This paper presents a software synthesis method for hard real-time systems with energy constraints, taking into account DVS, inter-task relations and runtime overheads. The proposed method adopts a formal model based on time Petri nets in order to provide a basis for precise pre-runtime schedule generation.

Over the last decade, energy consumption has received considerable attention from the scientific community, which has developed several techniques for increasing energy savings in embedded systems. Nevertheless, considering time-critical systems, the adoption of energy saving techniques,such as dynamic voltage scaling, considerably hardens the system design, since stringent timing constraints may be affected. This work presents an environment for measuring and scheduling hard real-time systems with energy constraints, providing a set of integrated tools for automatic assistance in these design activities. Besides, a formal model based on time Petri nets is adopted in order to provide a basis for precise pre-runtime schedule generation as well as to allow property analysis and verification.

This work proposes the use of pre-defined generalized stochastic Petri net (GSPN) components to model and evaluate outbound logistics. The modeling process takes into account the automatic generation of these stochastic models from high-level end-user-oriented descriptions. The model resultant from this translation is guaranteed to have some expected Petri nets' properties. The Stochastic Logistics Optimizer Tool (SLOT) was implemented with the aim of supporting the modeling process based on a component library. Case studies were carried out and one of them is presented.

Interval methods have been applied to the outer estimation of solution sets for real interval linear systems of equations. This work presents the evaluation of results relating to a significant set of algorithms for solving interval linear systems of equations applied to the steady state solution of dependability models. These algorithms have been implemented using the Matlab Intlab framework. The ISPN analysis takes into account the effects of variabilities in exponential transition rates and immediate transition weights when calculating dependability indices. Case studies were conducted and an evaluation is presented comparing the algorithms which have been applied to solving steady state dependability studies.

The outbound logistics can determine the success or failure of an industry. It has a high participation in overall logistics costs and is a decisive factor for the Quality of Service (QoS). Measuring this activity and evaluating possible changes will help its management and can lead to a cost reduction and a better QoS. This paper proposes the use of predefined Generalized Stochastic Petri Nets (GSPN) components that allows the evaluation of vehicles utilization, storage levels and QoS indices. The model representing an outbound logistics scenario is obtained with a bottom-up approach, through the composition of these components, guaranteeing some expected GSPN properties. At the end of this paper, a study case made in a Brazilian meat processing industry is presented.

The globalized and highly competitive manufacturing industries have been imposing several challenges upon the field of production engineering. Tuning systems to high performance within cost constraints has been a reality and a critical factor of success for them. Hence, traditional performance metrics are not enough to supply these manufacturing industries. In order to improve the performance evaluation of such systems, this paper proposes a GSPN with energy consumption and human resources allocation annotations. The validation of the proposed model has been carried out through the execution of a case study inside a Brazilian meat processing industry.

Interval arithmetic has been applied to model uncertainties and variabilities. An interval model is a space, family or class of models in which there are parameters represented by intervals, instead of real numbers. This paper describes the interval generalized stochastic Petri net (IGSPN) as an interval extension to the GSPN model. The IGSPN analysis takes into account the effects of variability on exponential transition rates and weights when calculating dependability measures. IGSPN analysis may be useful as a tool for decisionmaking. A Case study is presented for describing the models as well as for presenting the analysis methods.

DVS (dynamic voltage scaling) has been adopted as an effective technique for reducing energy consumption in mobile embedded systems. Since reducing CPU supply voltage affects maximum CPU frequency, several concerns (e.g. deadline missing) may arise when adopting DVS in systems that contain hard real-time constraints. Although several scheduling approaches have been developed to address voltage scaling in time-critical systems, inter-task relations, such as precedence and exclusion relations, have been neglected. This work presents a method for hard real-time systems scheduling with DVS considering precedence and exclusion relations. The proposed work adopts a formal model based on time Petri nets in order to find a feasible schedule using a pre-runtime approach that satisfies timing and energy constraints.

Application servers have been widely adopted as distributed infrastructure (or middleware) for developing distributed systems. Current approaches for performance evaluation of application servers have mainly concentrated on the adoption of measurement techniques. This paper, however, focuses on the use of simulation techniques and presents an approach for performance modelling and evaluation of application servers using Petri nets. In order to illustrate how the proposed approach may be applied, Petri net models of JBoss application server are presented and their performance results are compared with ones that have been measured

In this study we focus on the specification and evaluation of parameterized generalized and stochastic Petri net (GSPN) models for reliability and safety estimates of embedded systems. The embedded system to be modeled is specified using the extended dependability block diagram (EDBD), a high-level system specification model, which is composed of several and diverse blocks: functional, decision, standby spare, multiple and subsystem. For each block, parameterized GSPN model parameters may be defined as: failure and repair rates, failure and success probabilities, redundancy and number of redundant components (if any), mean time to failure (MTTF) and mean time to repair (MTTR) among others. The parameterized solution helps the development of high-level automation tools. The parameterized GSPN models are concise and changeable models, based on few block models (with small variations). These models, depending on the parameters, can assume different structural configuration. Through Markovian and non-Markovian distributions functions, evaluation of series, parallel, m-out-of-n and other complex structures are possible. Non-Markovian distributions can be represented by composition of exponential distributions by means of method-of-stages through moment matching technique or by analytical expressions into a random switch, a GSPN construct. At the end, a case study related to a flight-control system solution is presented and estimates are shown which validate the proposed model.

Interval arithmetic has been applied to model uncertainties. Generalized stochastic Petri Net (GSPN) is one of most commonly applied timed Petri net models for performance and reliability evaluation. This paper describes the interval generalized stochastic Petri net (IGSPN) as an interval extension to the GSPN model. The IGSPN analysis takes into account the effects of variability on exponential transition rates and weights when calculating performance indices. Three case studies are presented for describing the model as well as for presenting the analysis methods. IGSPN analysis for performance evaluation may be useful for engineers and technicians as a tool for decision-making.

Sequential Control System in the industry has been used in applications based on Programmable Logical Controllers (PLC). These Systems are, in general, highly complex and with an operation cycle around 1ms or 10ms. PLC are, in general, expensive for theses high complex applications. In this work, a Dynamical Reconfigurable approach is presented, based on Xilinx Virtex-II FPGA architecture, operating as a virtual hardware machine. In this context, the control process is specified in the industrial standard language SFC/Petri net (Sequential Function Chart). For large controllers, a partial and dynamical reconfiguration mechanism takes place and the controller is split into multiple contexts, which are sequentially executed within the same FPGA, without violating the operation cycle of the system, in spite of the reconfiguration overhead. The solution is cost compatible with current PLC for complex applications and can reach better performance by exploration of the potential parallelism of control descriptions.

Dynamically Reconfigurable Systems (DRS) offer a very interesting alternative for embedded digital systems design. Tasks scheduling within a reconfigurable environment allows the development of systems with better execution performance, chip area economy and lower power consumption. This paper describes a Petri Net based methodology for the design of dynamically reconfigurable systems, where tasks scheduling has as prime objective the best temporal performance of the overall application. The methodology includes the generation of an embedded controller supporting the scheduling process in the target architecture.

The context of this work is related to embedded hard real-time systems development, more specifically, in the software generation phase. Embedded software has become much harder to design caused by the diversity of requirements and high complexity. Correctness and timeliness verification is an issue to be concerned. Usually, complex embedded real-time systems rely on specialized operating system kernels. However, operating systems may introduce significant overheads in execution time as well as in memory requirement. Software synthesis might be an alternative approach to operating systems usage, since it can generate tailored code for satisfying functional, performance, and resource constraints, and automatically generate runtime support (scheduling, resource management, communication, etc.) customized for each particular specification. However, the dispatcher and timer interrupt handler overheads are often neglect in software synthesis research. This paper provides a formal approach for system's modeling, and such model is adopted for synthesizing a timely and predictable scheduled code taking into account dispatcher and interrupt handler overheads.

This paper presents a methodology for mapping the live sequence chart (LSC) language to an equivalent coloured Petri net (CPN) model LSC is a language for system specification, allowing one to specify what should happen for all execution of the system as well as the modeling of anti-scenarios. However, analysis and verification of system properties are not possible. In order to allow diagnosis of inconsistent specifications, besides simulation, verification and analysis should be considered. Therefore, the proposition of a CPN model for LSC allows verification and analysis of system described in LSC, hence, contributing for increasing designers' confidence on the system development process and reducing risk that may lead to project failure.

This work presents a method for modeling and analysis of production systems with emphasis on inventory policy. Based on the structural analysis, it provides basic performance measures, such as throughput and cycle time. Due to this fact, the main goal of this work is to introduce a Petri net based approach to solve the mentioned problem for inventory control. Besides, a timed Petri net based approach for modeling is also presented.

This work aims to describe a methodology for scheduling and allocation of hardware contexts, in applications with high degree of parallelism, in a Run-Time-Reconfiguration (RTR) proceeding for a reconfigurable FPGA. The Scheduling approach is based on the hardware resource distribution in the FPGA architecture. The Scheduler is modeled as a Petri Net and the best performance yields the best scheduling. The hardware contexts allocation is based on a Left-Edge algorithm principle for rationalization of resources in scheduling approach. The adaptation of the algorithm considers that pre-located areas for loading of the contexts in the architecture are used.

This paper presents how to model and carry out the performance analysis of message-oriented middleware (MOM) using generalised stochastic Petri Net (GSPN) models. The results obtained from the Petri net analysis are compared against ones measured in a commercial MOM. Additionally, some results are presented in order to demonstrate the flexibility and the benefits of the proposed model. This research also focuses on how to improve the MOM performance by suggesting appropriated techniques and adjustments to the MOM basic architecture. Finally, we point out some decisions usually taken by systems administrators that may have a major impact on the performance of MOM systems.

Embedded hard real-time systems have stringent timing constraints that must be satisfied for the correct functioning of the system. Hence all tasks must be finished before their deadlines. In addition, there are systems where energy is another constraint that must also be satisfied. In this paper, a pre-runtime scheduling algorithm is presented in order to find schedules satisfying both timing and energy constraints. The proposed approach uses state space exploration for finding pre-runtime schedules. However, the main problem with such methods is the space size, which can exponentially grow. This paper tackles this problem through a depth-first search method for generating a partial timed labeled transition system derived from the time Petri net model.

The functioning of the computer as a control component within a larger overall application, as in the embedded systems, may affect the application's integrity as well as people and equipment involved by the application. A computer like any physical system is subject to failure with consequences ranging from inconvenience to catastrophe. This paper proposes high level models for fault tolerant mechanisms, in special TMR and recovery block, based on deterministic and stochastic Petri net (DSPN). By means of the proposed models it is possible to perform preliminary reliability analysis and the obtained results might be considered in a co-design methodology. The proposed approach allows the modeler to calculate the reliability of a fault tolerant embedded system as a function of the failure rate. In this paper this feature is extended to allow for the determination of the reliability combining a range of failure rates.

An interface process generation methodology, based on Petri nets, is described for fast integrating point-to-point communicating modules. Formal basis of this methodology ease behavioral property-checking and consistent execution of the generated interface process. The exposed technique allows fast incorporation of third-party cores into SoPC systems design where integration task is often a barrier for reusability.

One of the most intricate problems in the synthesis of real-time systems is the scheduling. We present a formal modeling methodology based on time Petri nets (TPN) and a framework for application-specific scheduler synthesis. Finding a feasible scheduling is not an easy task because this problem, in its general form, is NP-hard. The method proposed finds a scheduling, whether one exists, using state space exploration. The problem with this approach is the space size, which can be very large for medium to large systems. We show how to minimize this problem using behavior restrictions at system modeling, and a partial-order reduction method. Additionally, the algorithm proposed for finding a feasible schedule uses a depth-first search based method. Therefore, states are only generated if strictly necessary. It is verified through real-world experimental results that a schedule is found examining a reduced number of states.

This work presents a structural methodology for computing the number of storage units (registers) in hardware/software co-design context considering timing constraints. The proposed method is based on one intermediate model-data flow net, specified by our team, that takes into account timing precedence and data-dependency. The considered hardware/software co-design framework uses Petri nets as common formalism for performing quantitative and qualitative analysis.

Due to design constraints, many digital systems are implemented as mixed hardware and software components. An informal choice of implementing a part of the system, either in software or in hardware, can produce incorrect or unsatisfactory results. Co-design methodologies have been developed to aid the development of such systems. The partitioning phase is one of the tasks carried out by a co-design methodology. It divides the original system into components taking in account software or hardware implementation. This work presents a model to carry out data dependency analysis, which is one of the aspects considered for the partitioning algorithm of the PISH codesign methodology.

This work presents a method of hardware/software partitioning considering multiple software components. The proposed method uses Petri nets as a common formalism to perform quantitative and qualitative analysis. The use of Petri net permits one to use a specification nondependent partitioning method. As an intermediate format Petri net allows one to analyze properties of the specification and formally compute performance indices which are used in the partitioning process. This paper highlights methods of computing load balance, precedence relation degree and communication cost of behavioral description in order to perform the initial allocation along with partitioning. This paper is devoted to describing the initial allocation algorithm, although an overview of the general partitioning method is also presented.

This work presents a timed constraints capture method that uses Timed Petri nets for systems written in OCCAM in the context of Hardware/Software partitioning. We present a method for translation of occam programs into timed Petri-nets as well as a time analysis method in the context of hardware/software partitioning. The time analysis allows the capture of the execution time of processes and sub-process, which are used in the hardware/software partitioning process.