scholarly journals Optimizing Service Placement for Microservice Architecture in Clouds

2019 ◽  
Vol 9 (21) ◽  
pp. 4663
Author(s):  
Yang Hu ◽  
Cees de Laat ◽  
Zhiming Zhao
Keyword(s):  
Response Time ◽  
Resource Constraints ◽  
Performance Degradation ◽  
High Response ◽  
New Approach ◽  
Service Placement ◽  
Overall Performance ◽  
Microservice Architecture ◽  
Different Parts

As microservice architecture is becoming more popular than ever, developers intend to transform traditional monolithic applications into service-based applications (composed by a number of services). To deploy a service-based application in clouds, besides the resource demands of each service, the traffic demands between collaborative services are crucial for the overall performance. Poor handling of the traffic demands can result in severe performance degradation, such as high response time and jitter. However, current cluster schedulers fail to place services at the best possible machine, since they only consider the resource constraints but ignore the traffic demands between services. To address this problem, we propose a new approach to optimize the placement of service-based applications in clouds. The approach first partitions the application into several parts while keeping overall traffic between different parts to a minimum and then carefully packs the different parts into machines with respect to their resource demands and traffic demands. We implement a prototype scheduler and evaluate it with extensive experiments on testbed clusters. The results show that our approach outperforms existing container cluster schedulers and representative heuristics, leading to much less overall inter-machine traffic.

scholarly journals Deep Reinforcement Learning in Ice Hockey for Context-Aware Player Evaluation

2018 ◽  
Author(s):  
Guiliang Liu ◽  
Oliver Schulte
Keyword(s):  
Reinforcement Learning ◽  
Empirical Evaluation ◽  
Professional Sports ◽  
Ice Hockey ◽  
New Approach ◽  
The Neural Network ◽  
Overall Performance ◽  
Game Context ◽  
Player Performance ◽  
Q Function

A variety of machine learning models have been proposed to assess the performance of players in professional sports. However, they have only a limited ability to model how player performance depends on the game context. This paper proposes a new approach to capturing game context: we apply Deep Reinforcement Learning (DRL) to learn an action-value Q function from 3M play-by-play events in the National Hockey League (NHL). The neural network representation integrates both continuous context signals and game history, using a possession-based LSTM. The learned Q-function is used to value players' actions under different game contexts. To assess a player's overall performance, we introduce a novel Game Impact Metric (GIM) that aggregates the values of the player's actions. Empirical Evaluation shows GIM is consistent throughout a play season, and correlates highly with standard success measures and future salary.

BRIDGING VITAL SIGNS AND SOCIAL INTERACTIONS FOR RESOURCE-CONSTRAINED EPIDEMIC CONTROL

2013 ◽  
Vol 12 (03) ◽  
pp. 469-489 ◽  
Author(s):  
ZHAOYANG ZHANG ◽  
HONGGANG WANG ◽  
KEN C. K. LEE ◽  
HUA FANG
Keyword(s):  
Social Interactions ◽  
Resource Constraints ◽  
Vital Signs ◽  
Identification Algorithm ◽  
Epidemic Control ◽  
New Approach ◽  
Control Effectiveness ◽  
Critical Node Identification ◽  
Active Can ◽  
Epidemic Diseases

This paper proposes a new approach that uses people's social interaction behavior collected by mobile phones and vital signs collected by wireless body area networks (WBAN) for epidemic control. By this approach, infectious people who are socially active can be quickly identified to be quarantined. To realize this approach, we introduce a notion of critical network and critical node identification algorithm. Observing some resource constraints such as quarantine cost and hardware limitation, we focus on optimizing the proposed approach such that high epidemic control effectiveness is achieved while the corresponding overhead is minimized. Our simulation results demonstrate that our approach can effectively control the spread of epidemic diseases in various situations.

scholarly journals Multi-Objective Service Placement Scheme Based on Fuzzy-AHP System for Distributed Cloud Computing

2019 ◽  
Vol 9 (17) ◽  
pp. 3550 ◽  
Author(s):  
A-Young Son ◽  
Eui-Nam Huh
Keyword(s):  
Energy Efficiency ◽  
Cloud Computing ◽  
Response Time ◽  
Data Centers ◽  
Cloud Data ◽  
Multi Objective ◽  
Service Placement ◽  
Service Response Time ◽  
Cloud Data Centers ◽  
Distributed Cloud

With the rapid increase in the development of the cloud data centers, it is expected that massive data will be generated, which will decrease service response time for the cloud data centers. To improve the service response time, distributed cloud computing has been designed and researched for placement and migration from mobile devices close to edge servers that have secure resource computing. However, most of the related studies did not provide sufficient service efficiency for multi-objective factors such as energy efficiency, resource efficiency, and performance improvement. In addition, most of the existing approaches did not consider various metrics. Thus, to maximize energy efficiency, maximize performance, and reduce costs, we consider multi-metric factors by combining decision methods, according to user requirements. In order to satisfy the user’s requirements based on service, we propose an efficient service placement system named fuzzy- analytical hierarchical process and then analyze the metric that enables the decision and selection of a machine in the distributed cloud environment. Lastly, using different placement schemes, we demonstrate the performance of the proposed scheme.

scholarly journals Pollen Analysis of Glaciers in Special Relation to the Formation of Various Types of Glacier Bands

1949 ◽  
Vol 1 (06) ◽  
pp. 325-332 ◽  
Author(s):  
H. Godwin
Keyword(s):  
Pollen Analysis ◽  
Statistical Evidence ◽  
General Nature ◽  
Special Relation ◽  
New Approach ◽  
Flow Problems ◽  
Pollen Content ◽  
Glacier Ice ◽  
Glacier Flow ◽  
Different Parts

Abstract The author describes the work of Dr. V. Vareschi on the pollen content of glacier ice. A general account of the method of analysis is given. The results of statistical evidence of the analyses on several alpine glaciers are discussed, particularly in regard to the general nature of glacier flow in different parts of the glaciers. Vareschi’s suggestions as to the modes of origin of a certain type of ogive banding and of parallel banding (Pflugfurcheneis) are dealt with in the light of his findings. It is pointed out that what appears to be a valuable new approach to the solution of glacier flow problems needs elaboration before it can be accepted in all its bearings.

Master Plans, a New Approach for Contingency Planning in the Colombian Oil Industryi

2003 ◽  
Vol 2003 (1) ◽  
pp. 1189-1198
Author(s):  
Darío Miranda ◽  
Ana María Betancur ◽  
Guillermo Gutiérrez
Keyword(s):  
Emergency Response ◽  
Oil Spills ◽  
Oil Industry ◽  
Contingency Planning ◽  
Disaster Prevention ◽  
New Approach ◽  
Master Plans ◽  
Contingency Plans ◽  
Different Types ◽  
Overall Performance

ABSTRACT Up to a few years ago, contingency planning in the Colombian oil industry treated contingency plans for oil spills, fires and explosions as independent contingency plans. Strategies, operational procedures and equipment were selected for each type of emergency response. This condition reduced overall performance for different types of emergencies and restricted an emergency brigade's responsibility and commitment when facing an emergency. The members of a fire brigade, for example, did not have enough experience to control an oil spill and simply felt that they were not responsible for such work. In 1989, new legislation in Colombia established an approach for emergency response in which local, regional and national committees were created at municipal, departmental and national levels. These committees were responsible for emergency response coordination that depended on the emergency type and size, and on the resources involved. It soon became evident that individual contingency plans should be incorporated into a single comprehensive document. However, there were no guidelines available to provide contingency planners with the tools necessary to develop such integrated plans. The Colombian Petroleum Institute's new approach for designing and structuring Contingency Plans for the Colombian oil industry is presented in this document. New schemes for comprehensive emergency response are discussed, considering strategic, operational and data processing issues. Master Contingency Plans provide Local Contingency Plan developers with suitable tools that will enable them to develop their own plans in a comprehensive way, while coordinating the plans with the Colombian National System for Disaster Prevention and Response considered in Colombian law.

A New Approach to Turbulence

1997 ◽  
Vol 12 (18) ◽  
pp. 3121-3152 ◽  
Author(s):  
V. M. Canuto ◽  
M. S. Dubovikov
Keyword(s):  
Present Model ◽  
Large Strain ◽  
General Structure ◽  
Nonlinear Interactions ◽  
Closed Set ◽  
New Approach ◽  
Overall Performance ◽  
Sweeping Effect ◽  
Shear Driven Flows ◽  
Kolmogorov Constant

We propose a closed set of dynamic equations to describe turbulence. The equations are the result of systematic and heuristic elements. Specifically, the UV part of the nonlinear interactions, represented by a dynamical viscosity, is computed for a stirring force of a particular nature. However, since the results exhibit a general structure, we suggest heuristically to extend them to arbitrary flows. Because of nonrenormalizable divergences, the IR part of the nonlinear interactions has constituted a serious problem. We suggest a heuristic model, the basic ingredient of which is that the transfer of energy among eddies is mostly a local process. We show that possible adjustable parameters are actually fixed by the model itself. Because of the heuristic nature of one part of the model, its overall validity rests largely on the accumulated evidence gathered from checking its predictions against data from a wide variety of flows. The model has been tested against more than seventy turbulence statistics for homogeneous isotropic and anisotropic flows (the Kolmogorov constant is predicted to be Ko = 5/3). The overall performance is good. Here, we first extend the model to inhomogeneous flows and test the predictions using the newest laboratory and DNS data on turbulent convection at large Ra (Rayleigh number). The model reproduces both types of data quite accurately. Second, we study the problem of the so-called "sweeping effect" and derive the relation between the ω and k-spectra. Third, we show that for shear driven flows the present model reproduces well the data at large strain rates while the widely used K - ∊ model does not.

An Analysis of Engineers Information Seeking Activity

2013 ◽  
Author(s):  
Philip Cash ◽  
Tino Stanković ◽  
Mario Štorga
Keyword(s):  
Design Process ◽  
Information Seeking ◽  
Design Activity ◽  
Visualization Technique ◽  
New Approach ◽  
Engineering Design Process ◽  
Overall Performance ◽  
Types Of Information ◽  
Source Of Information ◽  
Visual Network

Information seeking is an important part of the engineering design process. In this context the Internet has become a significant source of information, shaping the way engineers work and interact. Current work has focused on characterizing this activity in terms of total time allocated to different information sources or overall tasks, such as evaluating. However, these approaches do not give an understanding of how engineers information seeking affects their problem solving activity and ultimately their performance in the context of the design process. As such, a new approach is needed to decompose the complexity of information seeking activity in order to more effectively support the evolving needs of engineering designers and design researchers. This paper addresses these issues by using an experimental study and network visualization technique to analyze Internet based information seeking activity and its affect on engineers overall performance during the information seeking/feasibility stage of the design process. The study uses both final year trainee engineers and practicing engineers in order to more fully explore the different modes of information seeking activity. With the study complete, the visual network analysis is used to explore patterns of information seeking and other design activity. Based on the results, three clearly differentiated types of information seeking activity are identified and discussed.

Sign in / Sign up

Export Citation Format

Share Document