Resource Scheduling

The performance of data grids for data intensive, real-time applications is highly dependent on the data dissemination algorithm employed in the system. Motivated by this fact, this study first formally defines the real-time splittable data dissemination problem (RTS/DDP) where data transfer requests can be routed over multiple paths to maximize the number of data transfers to be completed before their deadlines. Since RTS/DDP is proved to be NP-hard, four different heuristic algorithms, namely kSP/ESMP, kSP/BSMP, kDP/ESMP, and kDP/BSMP are proposed. The performance of these heuristic algorithms is analyzed through an extensive set of data grid system simulation scenarios. The simulation results reveal that a performance increase up to 8 % as compared to a very competitive single path data dissemination algorithm is possible.

The aim of this paper is to provide a description of machine learning based scheduling approach for high-loaded distributed systems that have patterns of tasks/queries that occur recurrently in workflow. The core of this approach is to predict the future workflow of the system depending on previous tasks/queries using supervised learning. First of all, the workflow is analyzed using hierarchical clustering to reveal sets of tasks/queries. Revealed sets of tasks/queries then undergo restructuring to represent patterns of recurrent tasks/queries. Later these patterns become the object of the forecasting process performed using neural network. Information on predicted tasks/queries is used by the resource management system (RMS) to perform efficient schedule. To estimate the performance of the described method it was at first realized as a module of the simulation tool Alea that models the work of high-performance distributed systems and then compared with other state-of-the-art scheduling algorithms. The simulation was produced for two datasets: in one of the experiments the proposed method showed best results, and in the other it was inferior to just a single method, though it was much better than commonly used standard scheduling algorithms.

Results from disaster research suggest that methods for coordination between individual emergency responders and organizations should recognize the independence and autonomy of these actors. These actor features are key factors in effective adaptation and improvisation of response to emergency situations which are inherently uncertain. Autonomy and adaptability are also well-known aspects of a multi-agent system (MAS). In this paper we present two MAS strategies that can effectively handle aircraft deicing incidents. These MAS strategies help improve to prevent and reduce e.g. airplane delays at deicing stations due to changing weather conditions or incidents at the station, where aircraft agents adopting pre-made plans that would act on behalf of aircraft pilots or companies, would only create havoc. Herein each agent using its own decision mechanism deliberates about the uncertainty in the problem domain and the preferences (or priorities) of the agents. Furthermore, taking both these issues into account each proposed MAS strategy outperforms a naive first-come, first-served coordination strategy. The simulation results help pilots and companies taking decisions with respect to the scheduling of the aircraft for deicing when unexpected incidents occur: they provide insights in the impacts and means for robust selection of incident-specific strategies on e.g. deicing station delays of (individual) aircraft.

Results from disaster research suggest that methods for coordination between individual emergency responders and organizations should recognize the independence and autonomy of these actors. These actor features are key factors in effective adaptation and improvisation of response to emergency situations which are inherently uncertain. Autonomy and adaptability are also well-known aspects of a multi-agent system (MAS). In this paper we present two MAS strategies that can effectively handle aircraft deicing incidents. These MAS strategies help improve to prevent and reduce e.g. airplane delays at deicing stations due to changing weather conditions or incidents at the station, where aircraft agents adopting pre-made plans that would act on behalf of aircraft pilots or companies, would only create havoc. Herein each agent using its own decision mechanism deliberates about the uncertainty in the problem domain and the preferences (or priorities) of the agents. Furthermore, taking both these issues into account each proposed MAS strategy outperforms a naive first-come, first-served coordination strategy. The simulation results help pilots and companies taking decisions with respect to the scheduling of the aircraft for deicing when unexpected incidents occur: they provide insights in the impacts and means for robust selection of incident-specific strategies on e.g. deicing station delays of (individual) aircraft.

Cloud computing is a promising utility-based distributed computing environment in which resources (hardware/software) are offered as a service over the Internet on a pay per use basis. It involves elastic resource provisioning capabilities and hence charge only for those Cloud resources that are actually needed. However, true elasticity and cost-effectiveness in the pay-per-use Cloud business model has not been achieved yet. Most of the auto-scaling techniques in Cloud allow horizontal scaling by adding duplicate instances. They do not fully consider both; users' expected performance and pricing of the services. Since, Cloud platforms offer a plethora of server configurations at different pricing, the ability to scale by varied size instances can provide greater elasticity and hence significant improvements in cost and resource utilization. In this work, a heterogeneity-aware auto scaling algorithm that adapts to the workload changes while maintaining the Quality of Service (QoS) is proposed. The proposed method uses online profiling of Cloud resources and workload history to analytically estimate subsequent resource requirements. It then employs a heuristic to provision the smallest set of different size resource configurations in order to meet the QoS targets at reduced cost and improved resource utilization. Simulation based experimental evaluation on representative web and scientific workload patterns indicate that the proposed auto-scaling technique exhibits better performance as compared to the other contemporary horizontal scaling approaches.

Robot systems largely depend on embedded systems to operate. The interfaces of those embedded systems, e.g. motor controllers or laser scanners, are often vendor-specific and therefore require a component that translates from/to the Robot Operating System (ROS) Middleware interface. In this work we present an implementation and evaluation of a ROS Middleware client based on the Contiki operating systems, which is suitable for constrained embedded devices, like wireless sensor nodes. We show that in-buffer processing of ROS messages without relying on dynamic memory allocation is possible. That message contents can be accessed conveniently via well-known concepts of the C language (structs) with negligible processing overhead compared to a C++-based client. And that the message-passing middleware concept of ROS fits nicely in Contiki's event-based nature. Furthermore, in order for an environment enriched with wireless sensor network to help robots in navigating, understanding and manipulating environments a direct integration is mandatory.

Processor speeds continue to outpace the memory subsys- tem making it neccesary to proactively acquire and re- tain important data. Current applications have an ever increasing number of dynamically allocated data struc- tures and these data structures occupy large footprints. A large portion of dynamically allocated data is accessed through pointers in the form of recursive data structures. Loads accessing

In this paper, we propose Runahead Threads (RaT) as a valuable solution for both reducing resource contention and exploiting memory-level parallelism in Simultaneous Multithreaded (SMT) processors. Our technique converts a resource intensive memory-bound thread to a speculative light thread under long-latency blocking memory operations. These speculative threads prefetch data and instructions with minimal resources, reducing critical resource conflicts between threads. We compare an SMT architecture using RaT to both state-of-the-art static fetch policies and dynamic resource control policies. In terms of throughput and fairness, our results show that RaT performs better than any other policy. The proposed mechanism improves average throughput by 37% regarding previous static fetch policies and by 28% compared to previous dynamic resource scheduling mechanisms. RaT also improves fairness by 36% and 30% respectively. In addition, the proposed mechanism permits register file size reduction of up to 60% in a SMT processor without performance degradation.

a b s t r a c t Research associated with Big Data in the Cloud will be important topic over the next few years. The topic includes work on demonstrating architectures, applications, services, experiments and simulations in the Cloud to support the cases related to adoption of Big Data. A common approach to Big Data in the Cloud to allow better access, performance and efficiency when analysing and understanding the data is to deliver Everything as a Service. Organisations adopting Big Data this way find the boundaries between private clouds, public clouds and Internet of Things (IoT) can be very thin. Volume, variety, velocity, veracity and value are the major factors in Big Data systems but there are other challenges to be resolved. The papers of this special issue address a variety of issues and concerns in Big Data, including: searching and processing Big Data, implementing and modelling event and workflow systems, visualisation modelling and simulation and aspects of social media.

As the demand for cloud computing services escalates, energy consumption in cloud data centers has surged, raising environmental and operational concerns. Efficient task scheduling stands as a critical technique for reducing energy consumption while maintaining Quality of Service (QoS). This paper investigates energy-efficient task scheduling algorithms with an emphasis on green computing. We examine and compare various strategies from existing literature and propose a structured methodology to enhance resource utilization and reduce carbon footprint. Key performance metrics include energy savings, load balancing, and execution time.

IEEE 802.16 mesh mode defines three types of resource scheduling: coordinated centralized, coordinated distributed, and uncoordinated distributed. While the standard defines the required procedures and messages for each scheduler, it does not offer encouraging means to provide performance, reliability, or QoS. In this article we outline the issues of IEEE 802.16 mesh schedulers. We also survey representative proposals and qualitatively evaluate them against these issues. More critically, we identify key challenges that have not been addressed so far in the literature to motivate work in this area of research.

In this paper, we propose to use resource reservations scheduling and feedback-based allocation techniques for the provisioning of proper timeliness guarantees to audio processing applications. This allows real-time audio tasks to meet the tight timing constraints characterizing them, even if other interactive activities are present in the system. The JACK sound infrastructure has been modified, leveraging the real-time scheduler present in the Adaptive Quality of Service Architecture (AQuoSA). The effectiveness of the proposed approach, which does not require any modifiction to existing JACK clients, is validated through extensive experiments under different load conditions.

3GPP has introduced LTE Femtocells to manipulate the traffic for indoor users and to minimize the charge on the Macro cells. A key mechanism in the LTE traffic handling is the packet scheduler which is in charge of allocating resources to active flows in both the frequency and time dimension. So several scheduling algorithms need to be analyzed for femtocells networks. In this paper we introduce a performance analysis of three distinct scheduling algorithms of mixed type of traffic flows in LTE femtocells networks. The particularly study is evaluated in terms of throughput, packet loss ratio, fairness index and spectral efficiency.

A major method of constructing applications to run on a computational Grid is to assemble them from components -separately deployable units of computation of well-defined functionality. Performance steering is an adaptive process involving run-time adjustment of factors affecting the performance of an application. This paper presents a design for a system capable of steering, towards a minimum run-time, the performance of a component-based application executing in a distributed fashion on a computational Grid. The proposed performance steering system controls the performance of single applications, and the basic design seeks to separate application-level and component-level concerns. The existence of a middleware resource scheduler external to the performance steering system is assumed, and potential problems are discussed. A possible model of operation is given in terms of application and component execution phases. The need for performance prediction capability, and for repositories of application-specific and component-specific performance information, is discussed. An initial implementation is briefly described.

Cloud computing platforms must efficiently allocate compute resources to meet dynamic workload demands while minimizing cost. This paper proposes an adaptive resource allocation framework that uses workload forecasting and user-driven policies to optimize cloud resource usage in real time. By leveraging the data volume predictions and workload pattern analysis, our system proactively scales computing resources up or down in anticipation of needs, rather than relying solely on reactive auto-scaling. We mathematically formulate the resource allocation problem to optimize cost and performance, and we introduce a predictive allocation algorithm that adjusts virtual machine instances, memory, and storage based on forecasted job demands. The approach is compared against traditional static allocation and reactive scaling methods. Experimental results on a cloud testbed show that the predictive resource management strategy achieves up to 30% reduction in infrastructure cost and about 25% improvement in job execution times, without violating service-level agreements. We also incorporate a user-driven component allowing manual overrides (e.g., to respect budget limits or deadlines), adding flexibility. The paper includes performance evaluation graphs and tables demonstrating how adaptive provisioning outperforms

The use of asymmetric multi-core processors with onchip computational accelerators is becoming common in a variety of environments ranging from scientific computing to enterprise applications. The focus of current research has been on making efficient use of individual systems, and porting applications to asymmetric processors. In this paper, we take the next step by investigating the use of multi-corebased systems, especially the popular Cell processor, in a cluster setting. We present CellMR, an efficient and scalable implementation of the MapReduce framework for asymmetric Cell-based clusters. The novelty of CellMR lies in its adoption of a streaming approach to supporting MapReduce, and its adaptive resource scheduling schemes: Instead of allocating workloads to the components once, CellMR slices the input into small work units and streams them to the asymmetric nodes for efficient processing. Moreover, CellMR removes I/O bottlenecks by design, using a number of techniques, such as double-buffering and asynchronous I/O, to maximize cluster performance. Our evaluation of CellMR using typical MapReduce applications shows that it achieves 50.5% better performance compared to the standard nonstreaming approach, introduces a very small overhead on the manager irrespective of application input size, scales almost linearly with increasing number of compute nodes (a speedup of 6.9 on average, when using eight nodes compared to a single node), and adapts effectively the parameters of its resource management policy between applications with varying computation density.

Workflows generated from e-Science applications require the simultaneous use of multiple types of resources. In this paper, we present the Multiple Resource Reservation Model (M RRM ), which enables the monitoring and scheduling of multiple heterogeneous and distributed resources. M RRM provides a unified representation for multiple types of distributed resources, and represents resource constrains such as compatibility and accessibility. Using M RRM , we solve the Multiple Resource First Slot problem based on a collection of algorithms that are customized for different request and resource types. Our simulation results demonstrate efficiency and scalability of our algorithms for various sizes of networks and resource sets.

Fault-tolerant systems are crucial in environments where uninterrupted service is mandatory and also in environments where continuous service is critical, such as in cloud computing and high-availability systems. The main goal of this research is to introduce a fault-tolerant framework that can effectively manage faults without leading to system failures, ensuring seamless service delivery. The proposed system follows a two-phase approach, incorporating fault detection and correction mechanisms that are designed to maintain system integrity and minimize performance degradation. Upon identifying a fault, the system deploys a correction technique aimed at preventing Service Level Violations (SLV), which could otherwise disrupt the service and breach Service Level Agreements (SLA). This fault-handling process is designed to mitigate interruptions while enhancing overall system performance. The proposed solution demonstrates a significant improvement in reducing SLA violations, achieving a reduction to 98.76%. Additionally, the efficiency of the system is enhanced through an increase in the task success rate, further ensuring the reliability of service. To validate its effectiveness, the performance of this fault-tolerant system has been benchmarked against other contemporary algorithms such as First Come First Serve (FCFS), Priority-based, and CD algorithms. The results reveal that the proposed approach offers notable improvements in response times, reducing them by 23%, 16%, and 33.1% respectively, in comparison to the FCFS, Prioritybased, and CD algorithms. These results demonstrate the system's superior capability in handling faults and maintaining service quality, making it an ideal choice for environments where uninterrupted service is mandatory.

Previous research on runahead execution took it for granted as a prefetch-only technique. Even though the results of instructions independent of an L2 miss are correctly computed during runahead mode, previous approaches discarded those results instead of trying to utilize them in normal mode execution. This paper evaluates the effect of reusing the results of preexecuted instructions on performance. We find that, even with an ideal scheme, it is not worthwhile to reuse the results of preexecuted instructions. Our analysis provides insights into why result reuse does not provide significant performance improvement in runahead processors and concludes that runahead execution should be employed as a prefetching mechanism rather than a full-blown prefetching/result-reuse mechanism.

Previous research on runahead execution took it for granted as a prefetch-only technique. Even though the results of instructions independent of an L2 miss are correctly computed during runahead mode, previous approaches discarded those results instead of trying to utilize them in normal mode execution. This paper evaluates the effect of reusing the results of preexecuted instructions on performance. We find that, even with an ideal scheme, it is not worthwhile to reuse the results of preexecuted instructions. Our analysis provides insights into why result reuse does not provide significant performance improvement in runahead processors and concludes that runahead execution should be employed as a prefetching mechanism rather than a full-blown prefetching/result-reuse mechanism.

A typical grid application requires several processors for execution that may not be fulfilled by single cluster at times. Co-allocation is the concept of aggregating computing resources from more than one cluster to facilitate application execution. It poses great difficulty in implementing as these resources are distributed and managed locally. In this paper, we propose a metascheduling framework that achieves co-allocation using the concept of virtualization. Our approach differs from earlier ones as we create virtual machines to meet the requirements of application thereby utilizing the resources to the fullest possible extent while preserving their autonomy. We used Deviation Based Resource Scheduling algorithm to initiate SLA negotiation with other resources for participating in resource co-allocation. It also supports SLA monitoring and enforcement. Our preliminary results show that this approach achieves greater throughput against conventional scheduling.

Holocene variability in technology and subsistence economy in central Alaska is examined through a coarse-grained synthesis of dated components. Remarkable continuity during this period is evident in the ubiquity of microblade technology and the breadth of diet, including large and small mammals, birds, and fi sh. Signifi cant changes in technology can, however, be tracked around 6000 and 1000 cal B.P. 1 These seem to relate to changes in resource scheduling and land use strategies, shifts from residential to logistic mobility, and increased reliance on storage. Heuristic scenarios integrating these elements are offered to explain these transitions.

Efficient resource management in data centers and clouds running large distributed data processing frameworks like MapReduce is crucial for enhancing the performance of hosted applications and increasing resource utilization. However, existing resource scheduling schemes in Hadoop MapReduce allocate resources at the granularity of fixed-size, static portions of nodes, called slots. In this work, we show that MapReduce jobs have widely varying demands for multiple resources, making the static and fixed-size slot-level resource allocation a poor choice both from the performance and resource utilization standpoints. Furthermore, lack of coordination in the management of multiple resources across nodes prevents dynamic slot reconfiguration, and leads to resource contention. Motivated by this, we propose MROrchestrator, a MapReduce resource Orchestrator framework, which can dynamically identify resource bottlenecks, and resolve them through fine-grained, coordinated, and on-demand resource allocations. We have implemented MROrchestrator on two 24-node native and virtualized Hadoop clusters. Experimental results with a suite of representative MapReduce benchmarks demonstrate up to 38% reduction in job completion times, and up to 25% increase in resource utilization. We further demonstrate the performance boost in existing resource managers like NGM and Mesos, when augmented with MROrchestrator.

IT service providers employ server virtualization as a basic building block of their platforms to increase the cost effectiveness. The economic benefits from server virtualization come from higher resource utilization, reduced maintenance and operational costs including energy consumption. However, those benefits require efficient assignments of virtual servers or jobs to a limited number of physical hosts. The primary measures of evaluating efficiency of an assignment are that all the system resources are utilized effectively and the performance of each virtual machine is consistent with the desired performance bounds. While satisfying such SLA/performance bounds is essential for many classes of applications, the interference among the virtual machines makes such assurance admittedly difficult. This is primarily due to the complexity of estimating the completion time of jobs on assigned virtual servers. In this paper, we present a mathematical model for estimating the processing time of jobs in shared virtual servers by treating it as a conventional job scheduling problem. Given a job set, its total processing time is the fundamental objective function for a class of optimization problems, commonly known as job scheduling problems. The classical model that assumes the total processing time as the sum of individual processing times, unfortunately does not capture the multiple-resource contention to process a job. A job needs several resources such as processing cores and memory on the physical platform, hence the resource contention does not exhibit the simple classical behavior. In this work, we establish a general quantitative model for the total processing time for machines with multiple resources. A job is characterized by vector valued loading statistics and the total processing time of a job set is given by a quadratic function of those loading vectors. We validate the model with actual measurements of applications running times on a Xen-based virtualized infrastructure. The experimental results indicate that the proposed model is effective in capturing the resource contention in a virtualized platform. The average error rate of the estimated running time with synthetic workloads is 10.7% for identical jobs and 6.0% for heterogeneous jobs, and is within 6% with realistic workloads.

Buffer space management is critical to ensuring full utilization of network resources in present IoT networks. It allows the multiplexing of services with different quality of service requirements. Current IoT technologies supports a wide range of services including medical treatment, preventive equipment maintenance, remote operation of machinery, environmental monitoring, video surveillance and real-time alerts, connected transport as well as multiplexed services consisting of combinations of these. Traffic prioritization plays an important role in quality of service management in all mobile networks. From a network perspective, the aim is to ensure a minimum waiting time for packets of higher priority classes in the buffers of these networks. In order to use the resources correctly, it is necessary for each type of traffic or service to share the traffic capacity depending on the quality of service requirements they have.

The paper presents a framework for Virtual Laboratory (VLab). Authors focus their attention on the application structure, which can make it possible to remotely use very expensive devices. It is important that this structure is universal and can be used with every device and every simulation. Therefore the framework is a critical issue in that concept. To prove the concept

Specialized software, on-line tools and computational resources are very common in contemporary science. One of the exemplary domain is genomics-a new branch of science that developed rapidly in the last decade. As the genome research is very complex, it must be supported by professional informatics. In a microarray field the following steps cannot be performed without computational work: design of probes, quantitative analysis of hybridization results, post-processing, and finally data storage and management. Here, the general aspects of virtual laboratory systems are presented together with perspectives of their implementation in genomics in order to automate and facilitate this area of research.

In the user-interactive tasks the time slot reserved for running the task on a computational machine or scientific device must be synchronized with user preferences, considering specific work hours, daily schedule etc. In case of interactive experiments (eg. ...

In the paper some considerations about dynamic measurement scenarios aspects are discussed. The Dynamic Measurement Scenarios are used for creating and managing job workflow in the Virtual Laboratory system. It facilitate defining and monitoring measurement process which quite often consists of many tasks connected between themselves.

The paper presents a framework for Virtual Laboratory (VLab). Authors focus their attention on the application structure, which can make it possible to remotely use very expensive devices. It is important that this structure is universal and can be used with every device and every simulation. Therefore the framework is a critical issue in that concept. To prove the concept we plan to put into practice the Virtual Laboratory of Nuclear Magnetic Resonance Spectroscopy (NMR), which bases on this general structure. Another significant issue is to handle remote facilities like other Grid resources. This assumption generates new requirements concerning e.g. resource scheduling and global accounting.

We propose a simple model that is capable of capturing the statistical characteristics of mobile video traces encoded using MPEG4-Part2, AVC and SVC-TS encoding standards. The model can be adjusted to adapt to diverse workload configurations, which gives fellow researchers a great flexibility to evaluate different network traffic scenarios. We also discuss the modelbased trace generator and the challenges of its implementation. Moreover, we present the simulation results validating the model using different encoding settings. In addition to that, we provide several insights about our video modeling approach. This will help in testing, simulating and validating mobile video transmission and resource scheduling strategies over broadband wireless networks such as WiMAX and LTE..

Since the 1980's, efforts to restructure the electricity industry have received much attention around the world. To date, approximately twenty states have undergone restructuring to introduce formal markets as a vehicle for electricity transactions. Numerous other states are actively pursuing similar paths. Although the general principles which have guided restructuring are generally accepted, a variety of resulting utility and energy market structures have been implemented. In fact, opposing market philosophies have emerged. This paper reviews the present state of restructuring around the world, and describes key elements such as industry structures, market philosophies, dispatching strategies for spot markets and pricing.

This work is part of a project that aims to demonstrate the concept of Vehicle-to-Grid (V2G) with an operational fleet. A fleet of electric vehicles is operated with the objective of providing regulation services to the grid. The focus of this paper is on the real-time operation of the fleet. Specifically, given an optimal trajectory for the vehicle state of charge, schemes for distributing the regulation power commands among the vehicles are tested. A scheme based on a convex optimization problem is proposed. Several numerical illustrations and simulations show the effectiveness of the scheme respect to common scheduling heuristics in terms of accuracy.

In cloud computing, virtual containers on physical resources are provisioned to requesting users. Resource providers may pack as many containers as possible onto each of their physical machines, or may pack specific types and quantities of virtual containers based on user or system QoS objectives. Such elastic provisioning schemes for resource sharing may present major challenges to scientific parallel applications that require task synchronization during execution. Such elastic schemes may also inadvertently lower utilization of computing resources. In this paper, we describe the elasticity constraint effect and ripple effect that cause a negative impact to application response time and system utilization. We quantify the impact using real workload traces through simulation. Then, we demonstrate that some resource scheduling techniques can be effective in mitigating the impacts. We find that a tradeoff is needed among the elasticity of virtual containers, the complexity of scheduling algorithms, and the response time of applications.

In this paper, we analyze the behavior of several radio resource scheduling strategies under different power and subcarrier allocation (PSA) and inter-cell interference (ICI) coordination schemes in mobile OFDMA networks. The aim is to test the capability of these scheduling strategies to deal with quality of service (QoS) requirements both in a single-service scenario or when a mixture of services takes place. This work shows that radio resource allocation schemes managing ICI have strong impact on the scheduling policy behavior, and that traffic delay is a key parameter to track in order to provide QoS when services with different delay constraints simultaneously occur.

Lurette is an automated testing tool dedicated to reactive programs. The test process is automated at two levels: given a formal description of the System Under Test (SUT) environment, Lurette generates realistic input sequences; and, given a formal description of expected properties, Lurette performs the test results analysis. Lurette has been reimplemented from scratch. In this new version, the main novelty lies in the way the SUT environment is described. This is done by means of a new dedicated language called Lucky, a language for programming nondeterministic reactive machines. This article recalls the principles of Lurette, briefly presents the Lucky language, and describes some cases studies from the IST project Safeair II. The objective is to illustrate the usefulness of Lurette on real case studies, and the expressiveness of Lucky in accurately describing the SUT environment.

Cloud computing is a kind of distributed computing where different massively scalable IT related resources or capabilities are provided to a number of external users as a service using internet. Cloud computing is the delivery of computing services over the Internet. Cloud services allow duality and business to use software and hardware that are managed by third parties at remote locations. Here in this paper we have discussed many different load balancing techniques used to solve the issue in cloud computing and also discuss with quality of services in load balancing in cloud environment.

Adaptive, or self-aware, computing has been proposed to help application programmers confront the growing complexity of multicore software development. However, existing approaches to adaptive systems are largely ad hoc and often do not manage to incorporate the true performance goals of the applications they are designed to support. This paper presents an enabling technology for adaptive computing systems: Application Heartbeats. The Application Heartbeats framework provides a simple, standard programming interface that applications can use to indicate their performance and system software (and hardware) can use to query an application's performance. The PARSEC benchmark suite is instrumented with Application Heartbeats to show the broad applicability of the interface and an external resource scheduler demonstrates the use of the interface by assigning cores to an application to maintain a designated performance goal.

This paper describes an implementation of the 1-Step Ahead Minimum Inventory Variability Resource Scheduling Policy(R), in a large semiconductor facility (FAB) over the period from May, 1996, through January, 1997. The FAB described here uses a product release policy based on customer orders and a work-in-progress (WIP) chart. The scheduling of resource tools was done on a first in, first

The goal of resource leveling is to minimize the deviation between the resource requirements and the desired resource profile to prevent problems that caused by fluctuation of required resources. It is a common problem and arises frequently in real situations and therefore has been studied numerous times. Almost in all of these studies and the resulting solutions, there exist a common element, which is once an activity is started, it cannot be stopped and restarted again. That is, it cannot be split. In many instances in actual construction, there exist activities that can be split. So, it seems to be very useful to develop a model for resource leveling problems that allows certain activities to be split. An important interesting property that such a model should have is its simplicity. By simplicity we mean its ease of use by inexpert users without strong mathematical background. This paper presents a model based on genetic algorithms to level resources that permits selected activities to stop and restart. This splitting of activities results in improvement to the leveling solution that is traditionally achieved when splitting is not allowed. Examples are presented that illustrate the improvement in the solution obtained from the proposed model compared to models that do not allow splitting and compares the result to that obtained using commercially available software. The model presented here is very efficient and effective. Therefore, it can be used in real situations to automatically solve resource leveling problems efficiently to obtain effective results.

The cloud computing trend on the internet is vital as it allows data and applications to be managed over the internet instead of requiring personal devices. The job of users is scheduled in the resources of the cloud in order to improve performance. Scheduling tasks is an non-deterministic polynomial (NP)-hard problem, as it may have multiple solutions. Various researchers have proposed different load balancing and job scheduling algorithms to optimize the scheduling process in cloud environments, each with disadvantages. Therefore, this research proposes a novel hybrid load balancing and scheduling of tasks by the whale optimization algorithm (WOA) and seagull optimization algorithm (SOA) in the cloud. This hybrid proposed whale-seagull optimization algorithm (WSOA) optimizes task scheduling in the cloud by reducing processing time, response, and execution time, maximizing central processing unit (CPU) utilization, memory utilization, throughput, reliability, and balancing the load. The algorithm is simulated using the CloudSim toolkit package. As compared with existing approaches, simulation results showed better performance in terms of response time, processing time, execution time, CPU utilization, memory utilization, throughput, and reliability and is analyzed by comparing with the harries hawks optimization (HHO), hybrid dragonfly and firefly algorithm (ADA), spider monkey algorithm (SMA) and bird swarm optimization (BSO).

In this paper we describe the modelling of producer/consumer constraints with the CHIP system. Producer/consumer constraints arise in scheduling problems with consumable resources like raw materials or money, in particular for batch based processing. The constraint assures that at each time point enough consumable resources are available. The modelling with CHIP uses the cumulative constraint to express conditions in a very declarative way, yet obtains very good propagation due to the reasoning power build into the cumulative constraint. We show that with producer/consumer constraints many resource scheduling problems can be easily expressed and give examples of its industrial use.

h i g h l i g h t s • Presents a comprehensive survey of task scheduling strategies and the associated metrics suitable for cloud computing environments. • Discusses the various issues related to scheduling methodologies and the limitations to overcome. • Distinctive scheduling procedures are studied to discover which characteristics are to be included in a given system and which ones to disregard. • Literature survey organized based on three different perspectives: methods, applications, and parameter-based measures utilized. • Future research issues related to cloud computing-based scheduling identified.

Task execution in cloud computing requires obtaining stored data from remote data centers. Though this storage process reduces the memory constraints of the user’s computer, the time deadline is a serious concern. In this paper, Adaptive Cost-based Task Scheduling (ACTS) is proposed to provide data access to the virtual machines (VMs) within the deadline without increasing the cost. ACTS considers the data access completion time for selecting the cost effective path to access the data. To allocate data access paths, the data access completion time is computed by considering the mean and variance of the network service time and the arrival rate of network input/output requests. Then the task priority is assigned to the removed tasks based data access time. Finally, the cost of data paths are analyzed and allocated based on the task priority. Minimum cost path is allocated to the low priority tasks and fast access path are allocated to high priority tasks as to meet the time deadline....