Scientific Workflow Systems Research Articles

CMK: Enhancing Resource Usage Monitoring across Diverse Bioinformatics Workflow Management Systems

The increasing use of multiple Workflow Management Systems (WMS) employing various workflow languages and shared workflow repositories enhances the open-source bioinformatics ecosystem. Efficient resource utilization in these systems is crucial for keeping costs low and improving processing times, especially for large-scale bioinformatics workflows running in cloud environments. Recognizing this, our study introduces a novel reference architecture, Cloud Monitoring Kit (CMK), for a multi-platform monitoring system. Our solution is designed to generate uniform, aggregated metrics from containerized workflow tasks scheduled by different WMS. Central to the proposed solution is the use of task labeling methods, which enable convenient grouping and aggregating of metrics independent of the WMS employed. This approach builds upon existing technology, providing additional benefits of modularity and capacity to seamlessly integrate with other data processing or collection systems. We have developed and released an open-source implementation of our system, which we evaluated on Amazon Web Services (AWS) using a transcriptomics data analysis workflow executed on two scientific WMS. The findings of this study indicate that CMK provides valuable insights into resource utilization. In doing so, it paves the way for more efficient management of resources in containerized scientific workflows running in public cloud environments, and it provides a foundation for optimizing task configurations, reducing costs, and enhancing scheduling decisions. Overall, our solution addresses the immediate needs of bioinformatics workflows and offers a scalable and adaptable framework for future advancements in cloud-based scientific computing.

Hybrid fault tolerant cost aware mechanism for scientific workflow in cloud computing

Cloud computing provides solutions for diverse commercial and academic applications which is the primary goal. Scientific workflows are used in the cloud-computing environment to analyses large-scale scientific applications. For scientific workflows, many data is required, and a single scientific workflow that includes hundreds of stages, depending on the application's time restrictions, task failures, money limits, incorrect task organization, and task management issues can all hinder the implementation of scientific methods. In light of this, a cloud-based scientific workflow management and scheduling system that is fault-tolerant and data-oriented method are proposed. This research designs a novel hybrid cost-aware fault tolerant (HCFT) mechanism for minimizing the cost. Moreover, HCFT integrates optimal clustering and efficient resource utilization through parallel and distributed execution. Novelty of HCFT lies in novel clustering of the similar task for improvisation, CyberShake, laser interferometer gravitational wave observatory (LIGO), Montage, and sRNA identification protocol using high throughput technology (SIPHT) processes are used in the simulations to evaluate the performance of the proposed approach.

A cloud-based data processing and visualization pipeline for the fibre roll-out in Germany

To support the roll-out of fibre broadband Internet in Germany, Deutsche Telekom has set itself the goal of connecting more than 2.5 million households per year to FTTH (Fibre to the Home). However, planning and approval processes have been very complex and time-consuming in the past due to high communication overhead between stakeholders, missing automation, and lack of information about planning areas. Telekom addresses this problem by collecting large amounts of geospatial data (3D point clouds and 360°panorama images), which can be used to automatically find suitable routes for fibre optic lines, to determine possible locations for distribution cabinets, as well as to build a 3D visualization helping to create detailed plans and to present them to decision makers. This speeds up planning tremendously, but processing this data and creating the visualization in a short time requires automation. In this systems paper, we present a data processing platform that we have built and operated together with Telekom over the course of the last six and a half years. The platform makes use of the cloud to manage Big Data in a scalable and elastic manner. It builds upon research results from us, specifically a scientific workflow management system to automate processing, as well as Fibre3D, a web-based tool that planners can use to display the processed data and to perform fine-planning. Besides the technological aspects, this paper also describes a practical use case that shows how the platform and Fibre3D help Telekom speed up the planning and approval process. We also summarize lessons learned and give recommendations for the design of systems similar to ours. With the presented technology, Telekom has been able to already connect more than 8 million households to FTTH and expects to even improve on this in the future. We consider our collaboration, therefore, an example of how well knowledge and technology transfer between research and industry can work, and, at the same time, what impact it can have on society.

A qualitative assessment of using ChatGPT as large language model for scientific workflow development.

Scientific workflow systems are increasingly popular for expressing and executing complex data analysis pipelines over large datasets, as they offer reproducibility, dependability, and scalability of analyses by automatic parallelization on large compute clusters. However, implementing workflows is difficult due to the involvement of many black-box tools and the deep infrastructure stack necessary for their execution. Simultaneously, user-supporting tools are rare, and the number of available examples is much lower than in classical programming languages. To address these challenges, we investigate the efficiency of large language models (LLMs), specifically ChatGPT, to support users when dealing with scientific workflows. We performed 3 user studies in 2 scientific domains to evaluate ChatGPT for comprehending, adapting, and extending workflows. Our results indicate that LLMs efficiently interpret workflows but achieve lower performance for exchanging components or purposeful workflow extensions. We characterize their limitations in these challenging scenarios and suggest future research directions. Our results show a high accuracy for comprehending and explaining scientific workflows while achieving a reduced performance for modifying and extending workflow descriptions. These findings clearly illustrate the need for further research in this area.

Optimizing data regeneration and storage with data dependency for cloud scientific workflow systems

The characters of large volume, complex dependency and frequent reuse of the data in data-intensive scientific workflow systems make the data management more and more complex but crucial. Automating the data management is an urgent need for scientific workflow systems and, requires accurate models and efficient methods for data storage and request response. This paper deals with the problems of data regeneration and storage optimization by using data dependency for cloud scientific workflow systems. Through representing the data storage strategies as 0-1 strings, a global discrete model for data storage optimization is constructed, which has a complexity of O(2n) (n is the dataset number). In terms of data regeneration, we develop a deterministic method to determine optimal regeneration strategies for requested datasets with minimal computation under any data storage strategy, such that the systems can satisfy the data requests automatically and evaluate the computation cost accurately. To solve this storage optimization problem, the enumeration method and elitist canonical genetic algorithm are constructed by shortening the representation of storage strategies to avoid redundant computation. At last, experiments are conducted to test the proposed methods. Experimental results and comparisons show the positiveness and effectiveness of the proposed methods for data management in cloud scientific workflow systems.

Identification of RNA-binding Proteins in Spinal Cord Injury: An In-silico Approach

Introduction: Gene expression is regulated by trans-acting factors such as microRNA, and RNA-binding proteins (RBPs). Dysregulation of RNA-binding proteins (RBPs) are found in neurological diseases. However, the role RBPs in spinal cord injury (SCI) have not been identified. The objective of this study was to identify RBPs by re-analyzing RNA-sequencing data from SCI mice model using the latest version of Tuxedo pipeline. Methods: Reads from transcriptomic sequence of acute, subacute, and control mice models, from the Sequence Read Archive (SRA) website, were uploaded to a scientific workflow system called usegalaxy.org. The reads were assessed for their quality using FastQC, before they were mapped to the mouse mm10 reference genome using HISAT2. The fragments were then aligned to full-length transcripts using Stringtie, followed by DESeq2 to find differentially expressed genes (false discovery rate of 0.05 and fold change of -1< x >1). Finally, to find functional annotations, the Protein Analysis through Evolutionary Relationship (PANTHER) and g:Profiler were used. Results: There were 24 RBP-coding genes identified in the acute injury, and 27 in the subacute injury. Four RBPs that were commonly expressed at high levels in both acute and subacute injury; Hnrnpm, Ptbp3, Rbfox3 and Znf385a. These proteins regulate alternative splicing, and RNA transport. Other RBP-coding genes with a role in inflammatory response and apoptosis were also discovered. Conclusion: Novel RBP-coding genes differentially expressed in SCI were discovered, suggesting their role in the pathophysiology of SCI. These findings contribute to a better understanding of the regulatory mechanisms employed by RBPs in SCI.

K-span: Open and reproducible spatial analytics using scientific workflows

This paper describes the design, development, and testing of a general-purpose scientific-workflows tool for spatial analytics. Spatial analytics processes are frequently complex, both conceptually and computationally. Adaptation, documention, and reproduction of bespoke spatial analytics procedures represents a growing challenge today, particularly in this era of big spatial data. Scientific workflow systems hold the promise of increased openness and transparency with improved automation of spatial analytics processes. In this work, we built and implemented a KNIME spatial analytics (“K-span”) software tool, an extension to the general-purpose open-source KNIME scientific workflow platform. The tool augments KNIME with new spatial analytics nodes by linking to and integrating a range of existing open-source spatial software and libraries. The implementation of the K-span system is demonstrated and evaluated with a case study associated with the original process of construction of the Australian national DEM (Digital Elevation Model) in the Greater Brisbane area of Queensland, Australia by Geoscience Australia (GA). The outcomes of translating example spatial analytics process into a an open, transparent, documented, automated, and reproducible scientific workflow highlights the benefits of using our system and our general approach. These benefits may help in increasing users’ assurance and confidence in spatial data products and in understanding of the provenance of foundational spatial data sets across diverse uses and user groups.

Facilitating Asynchronous Collaboration in Scientific Workflow Composition Using Provenance

Advances in scientific domains are led to an increase in the complexity of the experiments. To address this growing complexity, scientists from different domains require to work collaboratively. Scientific Workflow Management Systems (SWfMSs) are popular tools for data-intensive experiments. To the best of our knowledge, very few of the existing SWfMSs support collaboration, and it is not efficient in many cases. Researchers share a single version of the workflow in existing collaborative data analysis systems, which increases the chance of interference as the number of collaborators grows. Moreover, for effective collaboration, contributors require a clear view of the project's status, the information that existing SWfMSs do not provide. Another significant problem is most scientists are not capable of adding collaborative tools to existing SWfMSs, and they need software engineers to take on this responsibility. Even for software engineers such tasks could be challenging and time consuming. In this paper, we attempted to address this crucial issue in scientific workflow composition and doing so in a collaborative setting. Hence, we propose a tool to facilitate collaborative workflow composition. This tool provides branching and versioning, which are standard version control system features to allow multiple researchers to contribute to the project asynchronously. We also suggest some visualizations and a variety of reports to increase group awareness and help the scientists to realize the project's status and issues. As a proof of concept, we developed an API to capture the provenance data and provide collaborative tools. This API is developed as an example for software engineers to help them understand how to integrate collaborative tools into any SWfMS. We collect provenance information during workflow composition and then employ it to track workflow versions using the proposed collaborative tool. Prior to implementing the visualizations, we surveyed to discover how much the proposed visualizations could contribute to group awareness. Moreover, in the survey we investigated to what extent the proposed version control system could help address shortcomings in collaborative experiments. The survey participants provided us with valuable feedback. In future, we will use the survey responses to enhance the proposed version control system and visualizations.

Open data and model integration through generic model agent toolkit in CyberWater framework

The CyberWater project is created to develop an open data and open model integration framework for studying complex environmental and water problems, where diverse online data sources can be directly accessed by diverse models without any need of users’ extra effort on the tedious tasks of data preparation for their models. We present our design and development of a novel generic model agent toolkit in the context of CyberWater, which enables users to integrate their models into the CyberWater system without writing any new code, significantly simplifying the data and model integration task. CyberWater adopts a visual scientific workflow system, VisTrails, which also supports provenance and reproducible computing. Our approach and the developed generic model agent toolkit are demonstrated, via CyberWater framework, with automated and flexible workflows through integrating data and models using real-world use cases. Two popular hydrological models, VIC and DHSVM, are used for illustrations.

Fault Tolerant and Data Oriented Scientific Workflows Management and Scheduling System in Cloud Computing

Cloud computing is a virtualized, scalable, ubiquitous, and distributed computing paradigm that provides resources and services dynamically in a subscription based environment. Cloud computing provides services through Cloud Service Providers (CSPs). Cloud computing is mainly used for delivering solutions to a large number of business and scientific applications. Large-scale scientific applications are evaluated through cloud computing in the form of scientific workflows. Scientific workflows are data-intensive applications, and a single scientific workflow may be comprised of thousands of tasks. Deadline constraints, task failures, budget constraints, improper organization and management of tasks can cause inconvenience in executing scientific workflows. Therefore, we proposed a fault-tolerant and data-oriented scientific workflow management and scheduling system (FD-SWMS) in cloud computing. The proposed strategy applies a multi-criteria-based approach to schedule and manage the tasks of scientific workflows. The proposed strategy considers the special characteristics of tasks in scientific workflows, i.e., the scientific workflow tasks are executed simultaneously in parallel, in pipelined, aggregated to form a single task, and distributed to create multiple tasks. The proposed strategy schedules the tasks based on the data-intensiveness, provides a fault tolerant technique through a cluster-based approach, and makes it energy efficient through a load sharing mechanism. In order to find the effectiveness of the proposed strategy, the simulations are carried out on WorkflowSim for Montage and CyberShake workflows. The proposed FD-SWMS strategy performs better as compared with the existing state-of-the-art strategies. The proposed strategy on average reduced execution time by 25%, 17%, 22%, and 16%, minimized the execution cost by 24%, 17%, 21%, and 16%, and decreased the energy consumption by 21%, 17%, 20%, and 16%, as compared with existing QFWMS, EDS-DC, CFD, and BDCWS strategies, respectively for Montage scientific workflow. Similarly, the proposed strategy on average reduced execution time by 48%, 17%, 25%, and 42%, minimized the execution cost by 45%, 11%, 16%, and 38%, and decreased the energy consumption by 27%, 25%, 32%, and 20%, as compared with existing QFWMS, EDS-DC, CFD, and BDCWS strategies, respectively for CyberShake scientific workflow.

Designing for Recommending Intermediate States in A Scientific Workflow Management System

To process a large amount of data sequentially and systematically, proper management of workflow components (i.e., modules, data, configurations, associations among ports and links) in a Scientific Workflow Management System (SWfMS) is inevitable. Managing data with provenance in a SWfMS to support reusability of workflows, modules, and data is not a simple task. Handling such components is even more burdensome for frequently assembled and executed complex workflows for investigating large datasets with different technologies (i.e., various learning algorithms or models). However, a great many studies propose various techniques and technologies for managing and recommending services in a SWfMS, but only a very few studies consider the management of data in a SWfMS for efficient storing and facilitating workflow executions. Furthermore, there is no study to inquire about the effectiveness and efficiency of such data management in a SWfMS from a user perspective. In this paper, we present and evaluate a GUI version of such a novel approach of intermediate data management with two use cases (Plant Phenotyping and Bioinformatics). The technique we call GUI-RISPTS (Recommending Intermediate States from Pipelines Considering Tool-States) can facilitate executions of workflows with processed data (i.e., intermediate outcomes of modules in a workflow) and can thus reduce the computational time of some modules in a SWfMS. We integrated GUI-RISPTS with an existing workflow management system called SciWorCS. In SciWorCS, we present an interface that users use for selecting the recommendation of intermediate states (i.e., modules' outcomes). We investigated GUI-RISPTS's effectiveness from users' perspectives along with measuring its overhead in terms of storage and efficiency in workflow execution.

Micro-Workflows Data Stream Processing Model for Industrial Internet of Things

The fog computing paradigm has become prominent in stream processing for IoT systems where cloud computing struggles from high latency challenges. It enables the deployment of computational resources between the edge and cloud layers and helps to resolve constraints, primarily due to the need to react in real-time to state changes, improve the locality of data storage, and overcome external communication channels’ limitations. There is an urgent need for tools and platforms to model, implement, manage, and monitor complex fog computing workflows. Traditional scientific workflow management systems (SWMSs) provide modularity and flexibility to design, execute, and monitor complex computational workflows used in smart industry applications. However, they are mainly focused on batch execution of jobs consisting of tightly coupled tasks. Integrating data streams into SWMSs of IoT systems is challenging. We proposed a microworkflow model to redesign the monolith architecture of workflow systems into a set of smaller and independent workflows that support stream processing. Micro-workflow is an independent data stream processing service that can be deployed on different layers of the fog computing environment. To validate the feasibility and practicability of the micro-workflow refactoring, we provide intensive experimental analysis evaluating the interval between sensor messages, the time interval required to create a message, between sending sensor message and receiving the message in SWMS, including data serialization, network latency, etc. We show that the proposed decoupling support of the independence of implementation, execution, development, maintenance, and cross-platform deployment, where each micro-workflow becomes a standalone computational unit, is a suitable mechanism for IoT stream processing.

An incremental reinforcement learning scheduling strategy for data‐intensive scientific workflows in the cloud

SummaryMost scientific experiments can be modeled as workflows. These workflows are usually computing‐ and data‐intensive, demanding the use of high‐performance computing environments such as clusters, grids, and clouds. This latter offers the advantage of the elasticity, which allows for changing the number of virtual machines (VMs) on demand. Workflows are typically managed using scientific workflow management systems (SWfMS). Many existing SWfMSs offer support for cloud‐based execution. Each SWfMS has its scheduler that follows a well‐defined cost function. However, such cost functions should consider the characteristics of a dynamic environment, such as live migrations or performance fluctuations, which are far from trivial to model. This article proposes a novel scheduling strategy, named ReASSIgN, based on reinforcement learning (RL). By relying on an RL technique, one may assume that there is an optimal (or suboptimal) solution for the scheduling problem, and aims at learning the best scheduling based on previous executions in the absence of a mathematical model of the environment. For this, an extension of a well‐known workflow simulator WorkflowSim is proposed to implement an RL strategy for scheduling workflows. Once the scheduling plan is generated via simulation, the workflow is executed in the cloud using SciCumulus SWfMS. We conducted a throughout evaluation of the proposed scheduling strategy using a real astronomy workflow named Montage.

INHIBITOR: An intrusion tolerant scheduling algorithm in cloud-based scientific workflow system

With the development of cloud computing technologies, more and more scientific workflows have been executed in clouds. However, cloud-based scientific workflows face many threats due to the resource sharing. Adversaries in clouds can directly or indirectly destroy them by means of side channels, virtual machine escape, and so on, which will cause interruption or produce incorrect outputs. Cloud-based scientific workflows are often applied in important scientific research fields, their failures will bring huge losses. Therefore, we propose an Intrusion toleraNt scHeduling algorIthm in cloud-Based scIenTific wORkflow system (INHIBITOR) to enhance the security. This algorithm constructs three replicas for each sub-task and designs a voting mechanism to realize the result verification. Based on this framework, INHIBITOR studies how to schedule these sub-task replicas and deduces the constraints which should be satisfied for intrusion tolerant scheduling. Furthermore, an elastic resource provisioning strategy is presented to improve resource utilization. To verify the effectiveness of INHIBITOR, we conduct experiments with WorkflowSim toolkit and use success rate, task completion rate and execution costs to evaluate it. Experimental results demonstrate that, compared with existing methods, INHIBITOR can not only increase the success rate by around 12.54%, but also improve the efficiency by about 4.7% and reduce execution costs by around 11.29%.

Solar radiation modeling with KNIME and Solar Analyst: Increasing environmental model reproducibility using scientific workflows

The inherent complexity of environmental models is frequently a limiting factor in their usefulness and practical applicability. This paper aims to demonstrate how scientific workflows can increase the reproducibility of environmental models by better managing this complexity. Specifically, through the example of Solar Analyst solar radiation model, the paper identifies three primary mechanisms for managing environmental modeling complexity using scientific workflows: (i) increasing transparency and improving reproducibility, in both the modeling process and the model itself; (ii) integrating validation and improving warrantability of solar radiation model outputs; and (iii) widening opportunities for supporting parameter-setting decisions for a diversity of modelers, using machine learning. The results demonstrate how each of these mechanisms can be realized using a freely-available and open-source scientific workflow management system (SWFMS) called KNIME. Firstly, our example KNIME workflows demonstrate increased transparency and improved reproducibility of solar radiation models and the entire modeling process. In turn, improving transparency and reproducibility can aid novice users in understanding and reusing solar radiation models. Secondly, an extended KNIME workflow is used to integrate both modeling and validation into a single, transparent workflow. Lastly, using KNIME workflows facilitates integration with other decision-support tools and techniques, such as machine learning. Using decision trees, an extended solar radiation KNIME workflow offers the capability to support more transparent and warrantable decisions around setting Solar Analyst parameter values. Ultimately, better managing the complexity of environmental modeling contributes to wider uptake and scrutiny of environmental models and the outputs they generate, both in scientific research and in applied evidence-based decision making.

VizSciFlow: A Visually Guided Scripting Framework for Supporting Complex Scientific Data Analysis

Scientific workflow management systems such as Galaxy, Taverna and Workspace, have been developed to automate scientific workflow management and are increasingly being used to accelerate the specification, execution, visualization, and monitoring of data-intensive tasks. For example, the popular bioinformatics platform Galaxy is installed on over 168 servers around the world and the social networking space myExperiment shares almost 4,000 Galaxy scientific workflows among its 10,665 members. Most of these systems offer graphical interfaces for composing workflows. However, while graphical languages are considered easier to use, graphical workflow models are more difficult to comprehend and maintain as they become larger and more complex. Text-based languages are considered harder to use but have the potential to provide a clean and concise expression of workflow even for large and complex workflows. A recent study showed that some scientists prefer script/text-based environments to perform complex scientific analysis with workflows. Unfortunately, such environments are unable to meet the needs of scientists who prefer graphical workflows. In order to address the needs of both types of scientists and at the same time to have script-based workflow models because of their underlying benefits, we propose a visually guided workflow modeling framework that combines interactive graphical user interface elements in an integrated development environment with the power of a domain-specific language to compose independently developed and loosely coupled services into workflows. Our domain-specific language provides scientists with a clean, concise, and abstract view of workflow to better support workflow modeling. As a proof of concept, we developed VizSciFlow, a generalized scientific workflow management system that can be customized for use in a variety of scientific domains. As a first use case, we configured and customized VizSciFlow for the bioinformatics domain. We conducted three user studies to assess its usability, expressiveness, efficiency, and flexibility. Results are promising, and in particular, our user studies show that VizSciFlow is more desirable for users to use than either Python or Galaxy for solving complex scientific problems.

A cost saving and load balancing task scheduling model for computational biology in heterogeneous cloud datacenters

Cloud-based scientific workflow systems can play an important role in the development of cost-effective bioinformatics analysis applications. There are differences in the cost control and performance of many kinds of servers in heterogeneous cloud data centers for bioinformatics workflows running, which can lead to imbalance between operational/maintenance management costs and quality of service of server clusters. A task scheduling model that responds to the peaks and valleys of task sequencing—the number of tasks that arrive in a given unit of time—is related to indicators such as cost saving, load balancing and system performance (average task wait time, average response time and throughput). This study proposes a large-scale cost-saving and load-balancing scheduling model, called HDCBS, for the optimization of system throughput. First, queuing theory is used to model each computing node as an independent queuing system and to obtain the average system wait time and average task response time. Then, using convex optimization theory, a task assignment solution is proposed with a load-balancing mechanism. The validity of the task scheduling model is verified by simulation experiments, and the model performance is further validated through a comparison with other frequently used scheduling methods. The simulation results show that the credibility of HDCBS is greater than 95% in task scheduling.

Protecting scientific workflows in clouds with an intrusion tolerant system

With the development of cloud computing technology, more and more scientific workflows are delivered to cloud platforms to complete. However, there are many threats in clouds due to the multi-tenant coexistence. In order to protect scientific workflows in clouds, the authors propose an intrusion tolerant scientific workflow system. In this system, the task executors containing multiple virtual machines are used for workflow sub-task execution to enhance reliability. Then lagged decision mechanism is presented to ensure uninterrupted workflow execution while checking the intermediate data, and assessing the confidence of these data. Inspired by moving target defence, they propose a dynamic task scheduling strategy based on resource circulation to periodically generate and recycle task executors, keeping the clean state of the workflow execution environment. Furthermore, temporary workflow intermediate data backup mechanism is presented, the stored intermediate data can be used for the re-execution of workflow sub-tasks with low confidence. Experiments are conducted in both the actual test environment based on OpenStack and the simulated test environment based on WorkflowSim toolkit. Experimental results demonstrate that the proposed system can effectively enhance intrusion tolerance of scientific workflows.

An open-data open-model framework for hydrological models’ integration, evaluation and application

To tackle fundamental scientific questions regarding health, resilience and sustainability of water resources which encompass multiple disciplines, researchers need to be able to easily access diverse data sources and to also effectively incorporate these data into heterogeneous models. To address these cyberinfrastructure challenges, a new sustainable and easy-to-use Open Data and Open Modeling framework called Meta-Scientific-Modeling (MSM) is developed. MSM addresses the challenges of accessing heterogeneous data sources via the Open Data architecture which facilitates integration of various external data sources. Data Agents are used to handle remote data access protocols, metadata standards, and source-specific implementations. The Open Modeling architecture allows different models to be easily integrated into MSM via Model Agents, enabling direct heterogeneous model coupling. MSM adopts a graphical scientific workflow system (VisTrails) and does not require re-compiling or adding interface codes for any diverse model integration. A study case is presented to illustrate the merit of MSM.

Makespan–Cost–Reliability-Optimized Workflow Scheduling Using Evolutionary Techniques in Clouds

Most popular scientific workflow systems can now support the deployment of tasks to the cloud. The execution of workflow on cloud has become a multi-objective scheduling in order to meet the needs of users in many aspects. Cost and makespan are considered to be the two most important objects. In addition to these, there are some other Quality-of-Service (QoS) parameters including system reliability, energy consumption and so on. Here, we focus on three objectives: cost, makespan and system reliability. In this paper, we propose a Multi-objective Evolutionary Algorithm on the Cloud (MEAC). In the algorithm, we design some novel schemes including problem-specific encoding and also evolutionary operations, such as crossover and mutation. Simulations on real-world and random workflows are conducted and the results show that MEAC can get on average about 5% higher hypervolume value than some other workflow scheduling algorithms.