Cloud Scheduling Research Articles

Efficient Task Scheduling for Quality of Service in Cloud Computing Network

Task scheduling in cloud is the process of allocating a resource to a task at specific time. The allocation of limited cloud resources to large number of tasks to satisfy the required quality of service is the key challenge in cloud. Allocation of a resource with less capability to a task increases the response time, makespan of the task and waiting time of the entire tasks in the waiting queue. This problem will result to an unsatisfied Quality of Service. In this paper we proposed an efficient task scheduling that uses three threshold values to specify the resource to be allocated to a task at a given time. This method ensures that a capable resource is allocated to task such that the response time and makespan of the all task are minimized. The proposed method was simulated using CloudSim and the result shows a better response time and makespan than the well known Min-Min and Max-Min Method.

RETRACTED: Reinforcement learning-based controller for adaptive workflow scheduling in multi-tenant cloud computing

Multi-tenancy is an essential feature in cloud computing and is a major component to achieve scalability and energy-efficient solution to gain high level of economic benefits. As the cloud, computing is gaining more audiences and high user base, the problem of scheduling the computational workflow for multi-tenant cloud scheduling is becoming a difficult task to achieve. In this study, we present a learning-based scheduler for catering heterogeneous software and hardware resources in the context of multi-tenant cloud computing. The experimentation has been carried out with the help of green cloud simulator and the results are compared with the state of the art techniques like minimum completion time, first come first serve and backfilling. The experimental results exhibit that the presented algorithm provides an effective means of utilizing cloud resources in addition with drastic reduction in cost of operation.

Scheduling mechanisms in the cloud environment: a methodological analysis

PurposeToday, with the rapid growth of cloud computing (CC), there exist several users that require to execute their tasks by the available resources to obtain the best performance, reduce response time and use resources. However, despite the significance of the scheduling issue in CC, as far as the authors know, there is not any systematic and inclusive paper about studying and analyzing the recent methods. This paper aims to review the current mechanisms and techniques, which can be addressed in this area.Design/methodology/approachThe central purpose of this paper refers to offering a complete study of the state-of-the-art planning algorithms in the cloud and also instructions for future research. Besides, this paper offers a methodological analysis of the scheduling mechanisms in the cloud environment.FindingsThe central role of this paper is to present a summary of the present issues related to scheduling in the cloud environment, providing a structure of some popular techniques in cloud scheduling scope and defining key areas for the development of cloud scheduling techniques in the future research.Research limitations/implicationsIn this paper, scheduling mechanisms are classified into two main categories include deterministic and non-deterministic algorithms; however, it can also be classified into different categories. In addition, the selection of all related papers could not be ensured. It is possible that some appropriate and related papers were removed in the search process.Practical implicationsAccording to the results of this paper, the requirement for more suitable algorithms exists to allocate tasks for resources in cloud environments. In addition, some principal rules in cloud scheduling should be re-evaluated to achieve maximum productivity and minimize wasted expense and effort. In this direction, to stay away from overloading and under loading of components and resources, the proposed method should execute workloads in an adaptable and scalable way. As the number of users increased in cloud environments, the number of tasks in the cloud that needed to be scheduled proportionally increased. Thus, an efficient mechanism is needed for scheduling tasks in these environments.Originality/valueThe general information gathered in this study makes the researchers acquainted with the state-of-the-art scheduling area of the cloud. Entirely, the answers to the research questions summarized the main objective of scheduling, current challenges, mechanisms and methods in the cloud systems. The authors hope that the results of this paper lead researchers to present more efficient scheduling techniques in cloud systems.

RETRACTED ARTICLE: A new whale optimizer for workflow scheduling in cloud computing environment

Cloud computing environments enable real time applications on virtualized resources that can be provisioned dynamically. It is one of the efficient platform service which permits to enable the various applications based on cloud infrastructure. Nowadays workflow systems become an easy and efficient task for the development of scientific applications. Efficient workflow scheduling algorithms are employed to improve the resource utilization by enhancing the cloud computing performance and to meet the users’ requirements. Many scheduling algorithms have been proposed but they are not optimal to incorporate benefits of cloud computing. In this paper a new framework are introduced as whale optimizer algorithm (WOA) which mimics the social behaviour of humpback whales and aims to maximize the work completion for meeting QoS constraints such as deadline and budget. This proposed method outperforms well when compared with other techniques and measured in terms of makespan, deadline and it is applicable for real time applications.

Hybrid Invasive Weed Optimization with Tabu Search Algorithm for an Energy and Deadline Aware Scheduling in Cloud Computing

The current existing high flexibility, profitability, and potential have made cloud computing extremely popular among the companies. This is used for improving and applying resources in an efficient manner and optimize makespan of the tasks. Scheduling is easy while there are only a few tasks to complete with few resources. Contrastingly, at the time the users forward several demands to the environment of the cloud, there may be a need for optimally selecting and allocating resources for achieving the desired quality of service that makes scheduling challenging. In this work, using intelligent metaheuristic algorithms for processing the requests and tasks of users in energy-aware scheduling made for a deadline is proposed. Genetic Algorithm (GA) the evolutionary algorithm that is inspired by the natural process of selection and the evolution theory. The Invasive Weed Optimization (IWO) was yet another novel stochastic based on the population that was a derivative-free technique of optimization inspired by the growth of the weed plants. The TABU Search (TS) was a generalization technique of local search where the TABU list was used for preventing cycling and further generating the candidates of the neighborhood. A hybrid GA with the TS (GA-TS) with a hybrid IWO with TS (IWO-TS) has been proposed for the energy and deadline aware scheduling. The framework further offers optimization of energy and performance. The primary purpose of this algorithm has been to improve deadline and scheduling in cloud computing along with local as well as global search algorithms. This framework will offer optimization of performance and energy. The reason behind presenting this algorithm was improving both scheduling and deadline in cloud computing using both local and global algorithms and results proved the algorithm to have better results.

Resource Constrained Profit Optimization Method for Task Scheduling in Edge Cloud

Edge cloud is a cloud computing system built on edge infrastructure. Task scheduling optimization is the key technology to ensure the quality of service in edge cloud. However, the openness of the edge cloud environment challenges the load balancing and profit optimization of task scheduling. In this paper, we analyze the business process and optimization factors of task scheduling in edge cloud. First, we propose a resource constrained task scheduling profit optimization algorithm (RCTSPO), which consists of clustering preprocessing, classification, profit matrix construction and optimal scheduling strategy calculation. Clustering preprocessing gathers similar tasks into one class and perform a classification on the clustered tasks. Then construct the profit matrix for resource constrained task scheduling, and the optimal task scheduling strategy is obtained based on the constructed profit matrix. Second, Petri nets are used to construct the different components of edge cloud, such as resource, task, user request and virtual machine, thus forming the task scheduling model of edge cloud. Third, the properties of task scheduling model are verified by using the related theory and tools of Petri nets. Finally, several experiments are done to evaluate the proposed method, the simulation results show that the algorithm not only achieves the maximum profit, but also performs well in terms of time, reliability and load balancing of task scheduling.

Implementation and evaluation of a trust model with data integrity-based scheduling in cloud

Implementation and evaluation of a trust model with data integrity-based scheduling in cloud

DEEP LEARNING BASED LOAD BALANCING USING MULTIDIMENSIONAL QUEUING LOAD OPTIMIZATION ALGORITHM FOR CLOUD ENVIRONMENT.

Cloud computing becoming one of the most advanced and promising technologies in these days for information technology era. It has also helped to reduce the cost of small and medium enterprises based on cloud provider services. Resource scheduling with load balancing is one of the primary and most important goals of the cloud computing scheduling process. Resource scheduling in cloud is a non-deterministic problem and is responsible for assigning tasks to virtual machines (VMs) by a servers or service providers in a way that increases the resource utilization and performance, reduces response time, and keeps the whole system balanced. So in this paper, we presented a model deep learning based resource scheduling and load balancing using multidimensional queuing load optimization (MQLO) algorithm with the concept of for cloud environment Multidimensional Resource Scheduling and Queuing Network (MRSQN) is used to detect the overloaded server and migrate them to VMs. Here, ANN is used as deep learning concept as a classifier that helps to identify the overloaded or under loaded servers or VMs and balanced them based on their basis parameters such as CPU, memory and bandwidth. In particular, the proposed ANN-based MQLO algorithm has improved the response time as well success rate. The simulation results show that the proposed ANN-based MQLO algorithm has improved the response time compared to the existing algorithms in terms of Average Success Rate, Resource Scheduling Efficiency, Energy Consumption and Response Time.

Implementation and evaluation of a trust model with data integrity-based scheduling in cloud

To address the cloud challenges, data integrity plays a vital role and provides accuracy and consistence of the stored data without modified. This work suggests a trust based min-min and max-min heuristic scheduling algorithms as several unknown parties or enterprises provide various services. The min-min and max-min is based on the completion time and execution time only. Hence, the trust factor is added to the conventional min-min and max-min algorithm along with the completion and execution time to support security and integrity in cloud. The specific characteristics of security within the cloud environment are assured using trusted third party (TTP). A trust evaluation model is presented as a solution which makes use of security mesh inside which the required trust is maintained.

Self-Managed Block Storage Scheduling for OpenStack-based Cloud

Abstract Software-defined storage (SDS) enables cloud providers to abstract the storage resources from underlying physical hardware by making storage resources programmable. It enhances the scalability and management of storage resources for better efficiency. Software-defined block storage component in OpenStack is referred to as cinder. The storage scheduling algorithm in OpenStack cinder is inefficient and one dimensional as it considers the capacity of storage disks as the only parameter to schedule the request. This compromises QoS to the customer. This paper presents the design of a self-managed block storage scheduling model. Proposed algorithms considers the performance attributes like read and write IOPS of storage hosts and classifies hosts by generating a quality status using machine learning classifiers. Further, this status is used in conjunction with a best -fit algorithm to make a final scheduling decision. Private cloud test-bed with 5 cinder hosts is set up to test scheduler performance. Results clearly show that the proposed model outperforms the default scheduler by evenly distributing the volume requests across the cinder nodes. The proposed model also maintains the QoS by providing the best available storage backend for every volume request.

A multi-objective EBCO-TS algorithm for efficient task scheduling in mobile cloud computing

Cloud computing proposals a novel service to the users to permitting for consume the distant resources due to some faults faced in smart devices. This causes to the emergence of the newest computing model known as mobile cloud computing. It supports to the user that the mobile devices not require to execute the application in user device as a replacement of some jobs will be execute in distant cloud/cloudlet is named as offloading jobs which is the nearby to the device. At the similar time job scheduling is more critical commonly at both the mobile devices and the mobile cloud. To solving the task scheduling problem, we propose an EBCO-TS algorithm for priority based task scheduling with load balancing model. The proposed EBCO-TS consume the power only 85% compared to the existing methods and it shows better load balancing when the number of request increases in the cloudlet.

An Algorithm Incarnating Deep Integration of Hardware-Software Energy Regulation Principles for Heterogeneous Green Scheduling

Heterogeneous green scheduling in virtual cloud is an urgent need of human sustainable developments. However, on the one hand, there is still considerable space beyond reach of the hardware energy regulation mode; on the other hand, as the core of green software methods, meta-heuristics algorithms are still underperforming in heterogeneous scheduling, although with many achievements in homogeneous scheduling. In this paper, an efficient new meta-heuristics algorithm is presented (i.e., GHSA_di), including the co-evolutionary dynamics equation emphasizing on and taking advantage of the hardware energy-regulation principles. The experimental results show that compared with the other three meta-heuristic scheduling algorithms, GHSA_di algorithm has obvious advantages in overall performance, energy saving and scalability, for both data intensive and computing intensive instances.

A multi-objective EBCO-TS algorithm for efficient task scheduling in mobile cloud computing

Cloud computing proposals a novel service to the users to permitting for consume the distant resources due to some faults faced in smart devices. This causes to the emergence of the newest computing model known as mobile cloud computing. It supports to the user that the mobile devices not require to execute the application in user device as a replacement of some jobs will be execute in distant cloud/cloudlet is named as offloading jobs which is the nearby to the device. At the similar time job scheduling is more critical commonly at both the mobile devices and the mobile cloud. To solving the task scheduling problem, we propose an EBCO-TS algorithm for priority based task scheduling with load balancing model. The proposed EBCO-TS consume the power only 85% compared to the existing methods and it shows better load balancing when the number of request increases in the cloudlet.

Open interactive education algorithm based on cloud computing and big data

In order to improve the self-learning ability of cloud computing and scheduling ability of big data resources in open interactive education, an open interactive education algorithm based on cloud computing and big data is proposed. An information flow model for open education big data is constructed, and big data mining is conducted to an open interactive education platform through the association rules mining method based on parallel scheduling to extract semantic ontology information feature quantity of interactive education; spatial attribute clustering is performed in the cloud computing environment according to the feature extraction results, and big data information is scheduled through the multi-feature weight allocation method. Simulation results show that in open interactive education, this method can cause relatively good output performance of big data mining, relatively high accuracy of feature information clustering of open interactive education and relatively strong feature resolution and recognition ability of data output, which meets the educational resource scheduling and allocation requirements of open education.

Agent–based monitoring of the task scheduling in computational clouds

Agent–based monitoring of the task scheduling in computational clouds

FACO: a hybrid fuzzy ant colony optimization algorithm for virtual machine scheduling in high-performance cloud computing

High-performance cloud computing has recently become the focus of much interest. Extensive research has shown that scheduling and load balancing are among the key aspects of performance optimization. The allocation of a set of requests into a set of computing resources, which is considered as an NP-hard problem, aims to distribute efficiently the load within the cloud architecture. To resolve this problem, the last decade has seen a growing trend towards using hybrid approaches to combine the advantages of different algorithms. In this paper, we propose a hybrid fuzzy ant colony optimization algorithm (FACO) for virtual machine scheduling to guarantee high-efficiency in a cloud environment. The proposed fuzzy module evaluates historical information to calculate the pheromone value and select a suitable server while keeping an optimal computing time. The experimental work presented in this study provides one of the first investigations into how to choose the optimal parameters of ant colony optimization algorithms using the Taguchi experimental design. We have simulated the proposed algorithm through the Cloud Analyst and CloudSim simulators by applying different cloud configurations to evaluate the performance of the proposed algorithm. Our findings highlight how response time and processing time are improved compared to the Round Robin algorithm, Throttled algorithm and Equally Spread Current Execution Load algorithm, especially in the case of a high number of nodes. FACO algorithm could be applied to define efficient cloud architecture adapted to high-performance applications.

Modelling and simulation of security-aware task scheduling in cloud computing based on Blockchain technology

Although a lot of work has been done in the domain, tasks scheduling and resource allocation in cloud computing remain the challenging problems for both industry and academia. Security in scheduling in highly distributed computing environments is one of the most important criteria in the era of personalization of the cloud services. Blockchain became recently a promising technology for integration with the cloud clusters and improvement of the security of cloud transactions and access to data and application codes. In this paper, we developed a new model of the cloud scheduler based on the blockchain technology. Differently to the other similar models, we tried to offload the implementation of the blockchain modules. We developed a novel ’proof–of–schedule’ consensus algorithm (instead of ’proof–of–work’) and used the Stackelberg games for the improvement of the approval of the generated schedules. The developed model has been experimentally simulated and validated by using the new original cloud simulator. The proposed Blockchain Scheduler was also compared with other selected cloud schedulers. The experiments shows that the applied approach improved significantly the efficiency of prepared schedules, in most cases, simulator returns a schedule with better makespan than existing individual scheduling modules.

Multiobjective noncooperative game model for cost‐based task scheduling in cloud computing

SummaryTask scheduling is an important issue in Cloud Computing. Cloud Computing often covers a range of different kinds of nodes, which offers a complex environment. To handle this complexity, as while as to process task scheduling in different scenarios of Cloud Computing, many algorithms have been studied. In this paper, we propose a game theory–based algorithm to the Cloud Computing task scheduling problem on the principle of minimizing the cost in the Cloud Computing. The Cloud Computing task scheduling problem is treated as a noncooperative game. The experiment results demonstrate that the game‐based algorithm has a desirable capability to reduce the cost in the Cloud Computing.

Chemical reaction optimization (CRO) for cloud job scheduling

Cloud computing is an emerging technology that provides functions of traditional computing as services via the Internet. Cloud job scheduling mechanisms try to allocate cloud resources to users’ submitted jobs in an optimal way. Several metaheuristic algorithms are used to obtain the optimum scheduling solution, such as genetic algorithm, glowworm swarm optimization and cat swarm optimization. This study introduced an optimal metaheuristic job scheduling method using chemical reaction optimization (CRO). Chemical reaction optimization is a new metaheuristic algorithm inspired from the interactions of molecules to achieve the lowest energy state possible during a chemical reaction. CRO mimics molecules’ interaction in chemical reaction microscopically. The CRO mechanism simulates the interactions of chemical reaction molecules to find out the lowest energy value possible. The proposed mechanism represents the molecular structure of each molecule as a vector of integer values, each of which represents a feasible cloud job scheduling solution. The potential energy ( $$PE$$ ) of each molecule represents its fitness in the objective function that denotes cloud scheduling execution time. Simulation using CloudSim simulator is used to evaluate the proposed CRO mechanism. A comparison with glowworm swarm optimization, cat swarm optimization and first come first served scheduling mechanism is made. The results of simulation showed that the CRO scheduling method has the shortest execution time among all other scheduling mechanisms.

Workload-Adaptive Configuration Tuning for Hierarchical Cloud Schedulers

Cluster schedulers provide flexible resource sharing mechanism for best-effort cloud jobs, which occupy a majority in modern datacenters. Properly tuning a scheduler's configurations is the key to these jobs’ performance because it decides how to allocate resources among them. Today's cloud scheduling systems usually rely on cluster operators to set the configuration and thus overlook the potential performance improvement through optimally configuring the scheduler according to the heterogeneous and dynamic cloud workloads. In this paper, we introduce AdaptiveConfig, a run-time configurator for cluster schedulers that automatically adapts to the changing workload and resource status in two steps. First, a comparison approach estimates jobs’ performances under different configurations and diverse scheduling scenarios. The key idea here is to transform a scheduler's resource allocation mechanism and their variable influence factors (configurations, scheduling constraints, available resources, and workload status) into business rules and facts in a rule engine, thereby reasoning about these correlated factors in job performance comparison. Second, a workload-adaptive optimizer transforms the cluster-level searching of huge configuration space into an equivalent dynamic programming problem that can be efficiently solved at scale. We implement AdaptiveConfig on the popular YARN Capacity and Fair schedulers and demonstrate its effectiveness using real-world Facebook and Google workloads, i.e., successfully finding best configurations for most of scheduling scenarios and considerably reducing latencies by a factor of two with low optimization time.