Meta-heuristic based autoscaling of cloud-based parameter sweep experiments with unreliable virtual machines instances

Monge, David A.; Pacini, Elina; Mateos, Cristian; Garino, Carlos García

doi:10.1016/j.compeleceng.2017.12.007

Cited by 15 publications

(40 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Note that this black-box, embarrassingly-parallel scheme to run instances of a singlethread code is actually a very popular way of conducting simulation-based experiments among scientists and engineers [46]. Many of such simulations execute the same application code (e.g.…”

Section: Methodsmentioning

confidence: 99%

Reducing energy usage in resource-intensive Java-based scientific applications via micro-benchmark based code refactorings

et al. 2019

Self Cite

View full text Add to dashboard Cite

In-silico research has grown considerably. Today's scientific code involves long-running computer simulations and hence powerful computing infrastructures are needed. Traditionally, research in high-performance computing has focused on executing code as fast as possible, while energy has been recently recognized as another goal to consider. Yet, energy-driven research has mostly focused on the hardware and middleware layers, but few efforts target the application level, where many energy-aware optimizations are possible. We revisit a catalog of Java primitives commonly used in OO scientific programming, or micro-benchmarks, to identify energy-friendly versions of the same primitive. We then apply the micro-benchmarks to classical scientific application kernels and machine learning algorithms for both single-thread and multi-thread implementations on a server. Energy usage reductions at the micro-benchmark level are substantial, while for applications obtained reductions range from 3.90% to 99.18%.

show abstract

Section: Methodsmentioning

confidence: 99%

Reducing energy usage in resource-intensive Java-based scientific applications via micro-benchmark based code refactorings

et al. 2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…CyberShake CyberShake is a workflow used in the Southern California Earthquake Center (SCEC) 4 for characterizing hazards generated by earthquakes. CyberShake performs a Probabilistic Seismic Hazard Analysis (PSHA) over a geographic region.…”

Section: Study Casesmentioning

confidence: 99%

“…In other of our works [4] we addressed the multi-objective minimization problem aiming at reducing makespan, monetary cost and the potential impact of OOB errors as stated in the previous paragraphs. However, in such article we focused on a different type of scientific application called parameter sweep experiments (PSEs).…”

Section: Introductionmentioning

confidence: 99%

CMI: An online multi-objective genetic autoscaler for scientific and engineering workflows in cloud infrastructures with unreliable virtual machines

Monge

Pacini

Mateos

et al. 2020

Journal of Network and Computer Applications

Self Cite

View full text Add to dashboard Cite

Cloud Computing is becoming the leading paradigm for executing scientific and engineering workflows. The large-scale nature of the experiments they model and their variable workloads make clouds the ideal execution environment due to prompt and elastic access to huge amounts of computing resources. Autoscalers are middleware-level software components that allow scaling up and down the computing platform by acquiring or terminating virtual machines (VM) at the time that workflow's tasks are being scheduled. In this work we propose a novel online multi-objective autoscaler for workflows denominated Cloud Multi-objective Intelligence (CMI), that aims at the minimization of makespan, monetary cost and the potential impact of errors derived from unreliable VMs. In addition, this problem is subject to monetary budget constraints. CMI is responsible for periodically solving the autoscaling problems encountered along the execution of a workflow. Simulation experiments on four well-known workflows exhibit that CMI significantly outperforms a state-of-the-art autoscaler of similar characteristics called Spot Instances Aware Autoscaling (SIAA). These results convey a solid base for deepening in the study of other meta-heuristic methods for autoscaling workflow applications using cheap but unreliable infrastructures.

show abstract

“…The main objectives of the proposed optimization model are identified as follows [14][15][16][17][18][19][20][21]. a.…”

Section: Proposed Task Scheduling Algorithmmentioning

confidence: 99%

Optimization model for QoS based task scheduling in cloud computing environment

Potluri

Rao

2020

IJEECS

View full text Add to dashboard Cite

Shortest job first task scheduling algorithm allocates task based on the length of the task, i.e the task that will have small execution time will be scheduled first and the longer tasks will be executed later based on system availability. Min- Min algorithm will schedule short tasks parallel and long tasks will follow them. Short tasks will be executed until the system is free to schedule and execute longer tasks. Task Particle optimization model can be used for allocating the tasks in the network of cloud computing network by applying Quality of Service (QoS) to satisfy user’s needs. The tasks are categorized into different groups. Every one group contains the tasks with attributes (types of users and tasks, size and latency of the task). Once the task is allocated to a particular group, scheduler starts assigning these tasks to accessible services. The proposed optimization model includes Resource and load balancing Optimization, Non-linear objective function, Resource allocation model, Queuing Cost Model, Cloud cost estimation model and Task Particle optimization model for task scheduling in cloud computing environement. The main objectives identified are as follows. To propose an efficient task scheduling algorithm which maps the tasks to resources by using a dynamic load based distributed queue for dependent tasks so as to reduce cost, execution and tardiness time and to improve resource utilization and fault tolerance. To develop a multi-objective optimization based VM consolidation technique by considering the precedence of tasks, load balancing and fault tolerance and to aim for efficient resource allocation and performance of data center operations. To achieve a better migration performance model to efficiently model the requirements of memory, networking and task scheduling. To propose a QoS based resource allocation model using fitness function to optimize execution cost, execution time, energy consumption and task rejection ratio and to increase the throughput. QoS parameters such as reliability, availability, degree of imbalance, performance and SLA violation and response time for cloud services can be used to deliver better cloud services.

show abstract

Meta-heuristic based autoscaling of cloud-based parameter sweep experiments with unreliable virtual machines instances

Cited by 15 publications

References 16 publications

Reducing energy usage in resource-intensive Java-based scientific applications via micro-benchmark based code refactorings

Reducing energy usage in resource-intensive Java-based scientific applications via micro-benchmark based code refactorings

CMI: An online multi-objective genetic autoscaler for scientific and engineering workflows in cloud infrastructures with unreliable virtual machines

Optimization model for QoS based task scheduling in cloud computing environment

Contact Info

Product

Resources

About