Comparative Analysis of Scientific Workflow Scheduling Algorithms in Cloud Under Budget Constraint

Complex huge-scale scientific applications are simplified by workflow to execute in the cloud environment. The cloud is an emerging concept that effectively executes workflows, but it has a range of issues that must be addressed for it to progress. Workflow scheduling using a nature-inspired metaheuristic algorithm is a recent central theme in the cloud computing paradigm. It is an NP-complete problem that fascinates researchers to explore the optimum solution using swarm intelligence. This is a wide area where researchers work for a long time to find an optimum solution but due to the lack of actual research direction, their objectives become faint. Our systematic and extensive analysis of scheduling approaches involves recently high-cited metaheuristic algorithms like Genetic Algorithms (GA), Whale Search Algorithm (WSA), Ant Colony Optimization (ACO), Bat Algorithm, Artificial Bee Colony (ABC), Cuckoo Algorithm, Firefly Algorithm and Particle Swarm Optimization (PSO). Based on various parameters, we do not only classify them but also furnish a comprehensive striking comparison among them with the hope that our efforts will assist recent researchers to select an appropriate technique for further undiscovered issues. We also draw the attention of present researchers towards some open issues to dig out unexplored areas like energy consumption, reliability and security for considering them as future research work.

Cloud computing is becoming very common in recent years and is growing rapidly due to its attractive benefits and features such as resource pooling, accessibility, availability, scalability, reliability, cost saving, security, flexibility, on-demand services, pay-per-use services, use from anywhere, quality of service, resilience, etc. With this rapid growth of cloud computing, there may exist too many users that require services or need to execute their tasks simultaneously by resources provided by service providers. To get these services with the best performance, and minimum cost, response time, makespan, effective use of resources, etc. an intelligent and efficient task scheduling technique is required and considered as one of the main and essential issues in the cloud computing environment. It is necessary for allocating tasks to the proper cloud resources and optimizing the overall system performance. To this end, researchers put huge efforts to develop several classes of scheduling algorithms to be suitable for the various computing environments and to satisfy the needs of the various types of individuals and organizations. This research article provides a classification of proposed scheduling strategies and developed algorithms in cloud computing environment along with the evaluation of their performance. A comparison of the performance of these algorithms with existing ones is also given. Additionally, the future research work in the reviewed articles (if available) is also pointed out. This research work includes a review of 88 task scheduling algorithms in cloud computing environment distributed over the seven scheduling classes suggested in this study. Each article deals with a novel scheduling technique and the performance improvement it introduces compared with previously existing task scheduling algorithms. Keywords: Cloud computing, Task scheduling, Load balancing, Makespan, Energy-aware, Turnaround time, Response time, Cost of task, QoS, Multi-objective. DOI: 10.7176/IKM/12-5-03 Publication date: September 30 th 2022

Cloud computing is the delivery of various services via the Internet. Data storage, databases, networking, software etc are examples of these resources. Cloud is growing daily and encounters numerous challenges. Among the challenges one of them is scheduling. There are many types of scheduling, one of them is task scheduling. In cloud computing, task scheduling is a issue that reduces system performance. The task scheduling issue is one of the critical issues in the cloud computing because cloud performance is heavily reliant on it. An effective algorithm is needed to increase system performance. In this paper, we attempt to study the performance of a task scheduling algorithm in a cloud data centre.

Infrastructure as a service model of cloud computing provides a tremendous amount of high-performance computing systems for the execution of scientific workflow applications. However, due to explosive growth in energy consumption and high charge cost of using these cloud systems, energy-efficient workflow scheduling under budget constraints becomes the most challenging issue. Very few research works have been done that consider the stated issue. Most of them mainly focus on the minimization of schedule length under user-specified budget constraints or energy consumption constraints. In this article, we propose an energy-efficient workflow scheduling algorithm named reducing energy consumption using fair pre-assignment of available budget (RECFPAB) that reduces energy consumption under client-specified budget constraints. The RECFPAB introduces a flexible mechanism to save energy consumption with the inclusion of energy and cost coefficient factor that enables fair distribution of available budget for unscheduled tasks of the workflow application. In order to compare the performance of the proposed algorithm, an energy-efficient version of the popular existing algorithms such as heterogeneous budget constrained scheduling and minimizing schedule length using budget level are introduced. The experimental evaluation based on Genome, LIGO, and Montage applications shows that RECFPAB gives significant results in comparison with considered algorithms.

Cloud computing recent emanation has drastically diversified everyone's perception of software delivery, infrastructure architectures, application development and deployment models. It is a new touchstone preceding the enterprise application development that can effectively expedite and facilitate the execution of various workflows in business process management system. Workflow is an automation of a business entity that can manage the business processes efficiently satisfying the requirements of prevailing enterprises. Dynamic workflows in a cloud computing environment empower the utilization of various cloud services to facilitate workflow execution. Scheduling multiple workflows in cloud environment is always a challenge. Workflow scheduling is a critical issue in the management of workflow execution. In this paper, we have compared and analyzed the performance of different heuristic workflow scheduling algorithms with various QoS parameters and scheduling factors.

Scheduling is a process that manages the workflow tasks during execution on different resources. Virtual infrastructure is a dynamic mapping of system resources to applications in order to maximize its utilization. In today's technological world, cloud has taken a long stride on the success towards maximum throughput as well as highest qualitative services to its consumers. Yet, approaches for maximizing the utilization of cloud resources are at peak demand. Each cloud service provider focuses on maximum utilization with minimum consumption of cloud resources, although managing and providing computational resources to maximum number of users and to execute such huge applications is a challenging one. In this paper, a scheduling algorithm with name TBW (Tabu Bayesian Whale Optimization) has been proposed. Basically, the algorithm is used to target the improvement in scheduling of scientific workflows. The complete framework has firstly used a ranking algorithm named distributed HEFT ranking and then applied TBW algorithm on ranked tasks of input workflows. The work has been executed for five scientific workflows LIGO, MONTAGE, Epigenomics, SIPHT and Cybershake. TBW is using tabu method on workflow tasks for fast local search in cloud system, and Bayesian Optimization is used to find out best possible combinations of resources where tasks are mapped and then whale optimization maps the tasks on the resources in a smart way. In the whole process, total execution time and cost parameters are minimized under deadline constraints.KeywordsCloud VMsTask mappingTabu searchBayesian optimizationWhale optimization

Scientific workflows benefit from the cloud computing paradigm, which offers access to virtual resources provisioned on pay-as-you-go and on-demand basis. Minimizing resources costs to meet user’s budget is very important in a cloud environment. Several optimization approaches have been proposed to improve the performance and the cost of data-intensive scientific Workflow Scheduling (DiSWS) in cloud computing. However, in the literature, the majority of the DiSWS approaches focused on the use of heuristic and metaheuristic as an optimization method. Furthermore, the tasks hierarchy in data-intensive scientific workflows has not been extensively explored in the current literature. Specifically, in this paper, a data-intensive scientific workflow is represented as a hierarchy, which specifies hierarchical relations between workflow tasks, and an approach for data-intensive workflow scheduling applications is proposed. In this approach, first, the datasets and workflow tasks are modeled as a conditional probability matrix (CPM). Second, several data transformation and hierarchical clustering are applied to the CPM structure to determine the minimum number of virtual machines needed for the workflow execution. In this approach, the hierarchical clustering is done with respect to the budget imposed by the user. After data transformation and hierarchical clustering, the amount of data transmitted between clusters can be reduced, which can improve cost and makespan of the workflow by optimizing the use of virtual resources and network bandwidth. The performance and cost are analyzed using an extension of Cloudsim simulation tool and compared with existing multi-objective approaches. The results demonstrate that our approach reduces resources cost with respect to the user budgets.

Workflow scheduling is one of the burning topics that has drawn enormous attention recently in the research community of cloud computing due to its wide applications in astronomy, physics, bioinformatics, health care and so on. This is a well-known NP-complete problem. It presents an interesting aspect of achieving minimum processing time of all the tasks and maximum resource utilization in cloud resources. Therefore, many algorithms have been developed for workflow scheduling. However, most of them consider a static priority of the tasks which is non-realistic for heterogeneous cloud computing environment. In this paper, we propose a workflow scheduling algorithm which considers dynamic priority of the tasks. The algorithm undergoes a process of min–max normalization followed by the calculation of the dynamic threshold to dispatch the tasks into one of the virtual machines. The algorithm is extensively simulated using various benchmark, scientific and real-life workflows. All the simulated results are compared with other four existing workflow scheduling algorithms. The simulated results confirm that the proposed algorithm lags behind all the four existing algorithms in terms of makespan and average cloud resource utilization. The simulation results are also validated through analysis of variance statistical test.

Scheduling scientific workflows in cloud environments is challenging due to the need to balance multiple objectives while satisfying various constraints. In this paper, we propose a Hybrid cloud Economic and Secure workflow scheduling system (HyEcoSec) designed to minimize makespan and cost while ensuring security and meeting budget and deadline constraints. HyEcoSec comprises two modules: a Security Compliance Module to manage user-defined security requirements, and a Scheduling Module integrating a static scheduler based on the Multi-Portions Slime Mould Algorithm (MPSMA) and a dynamic re-scheduler to handle runtime failures. MPSMA, an enhanced version of the Slime Mould Algorithm (SMA), balances makespan and cost using a modified Pareto approach under budget and deadline constraints. Performance evaluation using the WorkowSim demonstrates the effectiveness of HyEcoSec. Compared to baseline methods, MPSMA achieves a competitive execution time, reduces makespan by 26% and cost by 43%, and improves resource utilization by 22%. It ensures compliance with budget and deadline constraints in almost all cases, outperforming the compared algorithms. The integration of security services shows minimal cost and time impact, confirming HyEcoSec's suitability for secure environments. Furthermore, the dynamic re-scheduler enhances execution efficiency, reducing makespan by 61% and cost by 49% in failure scenarios.

Large-scale distributed machine learning (ML) systems involve extensive and costly computational resources. Pricing and scheduling, as two promising techniques for resource management, have garnered significant attention. However, existing job pricing and scheduling algorithms in cloud computing either charge fixed resource fees based on known job runtime or implement dynamic price setting with job preemption, unsuitable for distributed ML systems with high uncertainties and switching cost. First, whether the resources of a distributed ML job are placed together or not results in different job runtime. Second, various time-varying factors, including job arrival rates and competitors’ pricing, affect resource prices. Third, frequent price changes for the same resource can easily lead to system instability, ultimately jeopardizing user satisfaction. Addressing these uncertainties is challenging. This article introduces <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">APOS</monospace>, an adaptive pricing and online scheduling framework, aiming at maximizing the operator’s overall revenue. <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">APOS</monospace> incorporates two innovations: 1) Intelligent Pricing: We represent each price using a feature vector that encapsulates relevant factors. Subsequently, based on the linear upper confidence bound (UCB) techniques, we establish relationships between price features and two revenue-associated elements: a) job arrival rates and b) resource consumption rates. To ensure system stability, we introduce batch pricing to reduce the frequency of resource price updates and 2) Online Scheduling: We strive to compute a nonpreemptive schedule that balances job utility with corresponding resource cost. We rigorously prove that <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">APOS</monospace> achieves truthfulness, individual rationality, system stability, and sublinear regret in polynomial time. Finally, extensive trace-driven simulations confirm that <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">APOS</monospace> outperforms four state-of-the-art baselines, yielding a minimum of 23.3% improvement in total operator revenue.

Multi-objective optimization for scientific workflow scheduling based on Performance-to-Power Ratio in fog–cloud environments

Cost optimization approaches for scientific workflow scheduling in cloud and grid computing: A review, classifications, and open issues

Energy-efficient execution of the scientific workflow is a challenging task in cloud computing that demands high-performance computing to process growing datasets. Due to the interdependency of tasks in the scientific workflow applications, energy-efficient resource allocation is vital for large-scale applications running on heterogeneous physical machines. Thus, this paper proposes a hybrid heuristic algorithm based energy-efficient cloud computing service (HH-ECO) that offers a significant solution for resource allocation, task scheduling, and optimization of scientific workflows. To ensure the energy-efficient execution, the HH-ECO focuses on executing non-dominant workflow tasks through adaptive mutation and energy-aware migration strategy. HH-ECO adopts the chaotic based particle swarm optimization (C-PSO) principle to optimize the resource allocation, task scheduling, and resource migration by generating the global best plans without local convergence. C-PSO with adaptive mutation avoids the deterioration of global optima while finding the best host to place the virtual machine and ensures an appropriate resource allocation plan. By considering the workflow task precedence relationships during C-PSO based task scheduling, the novel hybrid heuristic method efficiently solves the multi-objective combinatorial optimization problem without dominance among the workflow tasks. The Cloudsim based simulation study delivers superior results compared to the existing methods such as the hybrid heuristic workflow scheduling algorithm (HHWS) and distributed dynamic VM management (DDVM). The proposed approach significantly improves the optimal makespan to 38.27% and energy conservation to 38.06% compared to the existing methods.

Minimizing the application running cost becomes a must in our day when Cloud Computing offers support for a large variety of applications. Being at the confluence with Cyber infrastructures and Big Data platforms, Cloud systems support efficient and effective workflows. Due to the increase in computational power for executing workflows, the need for efficient usage of said power also arises. After a brief overlook of current available workflow scheduling algorithms, such as HLFET - Highest Level First with Estimated Time or MCP - Modified Critical Path, it is clear that, for no specific constraints, they waste CPU resources, with no efficient task allocation. For this reason a new algorithm, called HER (Hybrid Algorithm for Workflow Scheduling in Cloud-based Cyber infrastructures) is proposed in this paper. HER algorithm reduces CPU idle time, with effect on the number of wasted resources and ensure a good load balancing for a cluster of resources available in a Cloud environment. The HER algorithm is exemplified on a number of different workflow scenarios in Cyber Water project - Prototype Cyber infrastructure based System for Decision-Making Support in Water Resources Management. The experimental approach considers several workflows such as decision support, real-time alerts, and visualization that use computational, storage and network resources in Clouds.

A heterogeneous system can be portrayed as a variety of unlike resources that can be locally or geologically spread, which is exploited to implement data-intensive and computationally intensive applications. The competence of implementing the scientific workflow applications on heterogeneous systems is determined by the approaches utilized to allocate the tasks to the proper resources. Cost and time necessity are evolving as different vital concerns of cloud computing environments such as data centers. In the area of scientific workflows, the difficulties of increased cost and time are highly challenging, as they elicit rigorous computational tasks over the communication network. For example, it was discovered that the time to execute a task in an unsuited resource consumes more cost and time in the cloud data centers. In this paper, a new cost- and time-efficient planning algorithm for scientific workflow scheduling has been proposed for heterogeneous systems in the cloud based upon the Predict Optimistic Time and Cost (POTC). The proposed algorithm computes the rank based not only on the completion time of the current task but also on the successor node in the critical path. Under a tight deadline, the running time of the workflow and the transfer cost are reduced by using this technique. The proposed approach is evaluated using true cases of data-exhaustive workflows compared with other algorithms from written works. The test result shows that our proposed method can remarkably decrease the cost and time of the experimented workflows while ensuring a better mapping of the task to the resource. In terms of makespan, speedup, and efficiency, the proposed algorithm surpasses the current existing algorithms—such as Endpoint communication contention-aware List Scheduling Heuristic (ELSH)), Predict Earliest Finish Time (PEFT), Budget-and Deadline-constrained heuristic-based upon HEFT (BDHEFT), Minimal Optimistic Processing Time (MOPT) and Predict Earlier Finish Time (PEFT)—while holding the same time complexity.