An Auto-Scaling Cloud Controller Using Fuzzy Q-Learning - Implementation in OpenStack

@inproceedings{Arabnejad2016AnAC,
  title={An Auto-Scaling Cloud Controller Using Fuzzy Q-Learning - Implementation in OpenStack},
  author={Hamid Arabnejad and Pooyan Jamshidi and Giovani Estrada and Nabil El Ioini and Claus Pahl},
  booktitle={ESOCC},
  year={2016}
}
Auto-scaling, i.e., acquiring and releasing resources automatically, is a central feature of cloud platforms. [] Key Method We propose a dynamic learning strategy based on a fuzzy logic algorithm, which learns and modifies fuzzy scaling rules at runtime without requiring prior knowledge.
A Comparison of Reinforcement Learning Techniques for Fuzzy Cloud Auto-Scaling
TLDR
Two dynamic learning strategies based on a fuzzy logic system, which learns and modifies fuzzy scaling rules at runtime, can handle various load traffic situations, sudden and periodic, and delivering resources on demand while reducing operating costs and preventing SLA violations.
RLPAS: Reinforcement Learning-based Proactive Auto-Scaler for Resource Provisioning in Cloud Environment
TLDR
The proposed RLPAS (Reinforcement Learning based Proactive Auto-Scaler) algorithm is based on the existing Reinforcement Learning (RL)-SARSA algorithm that learns the environment in parallel and allocates the resources and proves to be feasible for the extensive range of real cloud applications.
A Fuzzy Load Balancer for Adaptive Fault Tolerance Management in Cloud Platforms
TLDR
This work proposes a fuzzy job distributor (load balancer) for fault tolerance management to reduce levels of management complexity for the user and shows a significant reduction in the occurrence of faults in comparison with other load balancing algorithms.
NFV service dynamicity with a DevOps approach: Insights from a use-case realization
TLDR
This experience paper describes the process of leveraging the NFV orchestration platform built in the EU FP7 project UNIFY to deploy a dynamic network service exemplified by an elastic router, highlighting multiple learnings yielded throughout the prototype realization, focussing on the functional areas of service decomposition and scaling; programmable monitoring; and automated troubleshooting.
Scalability-aware mechanism based on workload prediction in ultra-peer networks
TLDR
This paper presents a scalability-aware approach for ultra-peers networks where each ultra-peer behaves like an ecosystem, in which, the presence of bottleneck is prevented in these structured information systems, and solutions to allow eachUltra-peer to scale with the growth of the network size are proposed.
Technical debt-aware elasticity management in cloud computing environments
TLDR
The results indicate that value-oriented adaptations in elasticity management lead to a better economics performance in terms of lower operating costs and higher quality of service over time.
MỘT MÔ HÌNH HỌC TĂNG CƯỜNG CHO VẤN ĐỀ ĐIỀU CHỈNH TỰ ĐỘNG TÀI NGUYÊN TRONG ĐIỆN TOÁN ĐÁM MÂY DỰA TRÊN FUZZY Q-LEARNING
TLDR
A model for auto-scaling resources based on fuzzy logic and reinforcement learning Fuzzy Q-Learning to increase adaptability as well as the ability to immediately respond to resource requirements in dynamic allocation/deallocation of resources in the direction of load balancing is built.
An Application-Oriented Autoscaler Based on Deep Learning
TLDR
An application-oriented autoscaler based on Long Short Term Memory network and Multi-Layer perceptron is presented, which includes a workload prediction model, a response time prediction model and a resource adjusting model that had better performance in workload prediction and server adjusting.
Applying Machine Learning in Self-adaptive Systems
TLDR
A systematic literature review of machine learning techniques in self-adaptive systems based on a traditional Monitor-Analyze-Plan-Execute (MAPE)-based feedback loop finds that machine learning is mostly used for updating adaptation rules and policies to improve system qualities, and managing resources to better balance qualities and resources.
A Cloud Broker for Service Adaptation Based on Interface Localization
TLDR
The objective is to support and implement an explicit modelling of aspects that are relevant to localization and runtime support consisting of tools and middleware services to automating the deployment based on models of locales, driven by the two localization dimensions.
...
1
2
...

References

SHOWING 1-10 OF 18 REFERENCES
Fuzzy Self-Learning Controllers for Elasticity Management in Dynamic Cloud Architectures
TLDR
FQL4KE is introduced, a self-learning fuzzy controller that learns and modifies fuzzy rules at runtime that empowers users to configure cloud controllers by simply adjusting weights representing priorities for architecture quality instead of defining complex rules.
Autonomic resource provisioning for cloud-based software
TLDR
F fuzzy logic is exploited to enable qualitative specification of elasticity rules for cloud-based software and a control theoretical approach using type-2 fuzzy logic systems to reason about elasticity under uncertainties is discussed.
A Review of Auto-scaling Techniques for Elastic Applications in Cloud Environments
TLDR
This work proposes a classification of techniques for automating application scaling in the cloud into five main categories: static threshold-based rules, control theory, reinforcement learning, queuing theory and time series analysis, and uses this classification to carry out a literature review of proposals.
Integrated and autonomic cloud resource scaling
TLDR
A Cloud resource auto-scaling system that addresses and overcomes above limitations and reduces the number of auto- scaling systems to be supported in a Cloud management system is described.
VCONF: a reinforcement learning approach to virtual machines auto-configuration
TLDR
A reinforcement learning (RL) based approach, namely VCONF, to automate the VM configuration process, which employs model-based RL algorithms to address the scalability and adaptability issues in applying RL in systems management.
An adaptive hybrid elasticity controller for cloud infrastructures
TLDR
This paper focuses on horizontal elasticity, the ability of the infrastructure to add or remove virtual machines allocated to a service deployed in the cloud, and builds two adaptive proactive controllers that estimate the future load on a service using queuing theory.
Lightweight Resource Scaling for Cloud Applications
  • Rui Han, Li Guo, M. Ghanem, Yike Guo
  • Computer Science
    2012 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (ccgrid 2012)
  • 2012
TLDR
This paper proposes a lightweight approach to enable cost-effective elasticity for cloud applications that operates fine-grained scaling at the resource level itself (CPUs, memory, I/O, etc) in addition to VM-level scaling.
From Data Center Resource Allocation to Control Theory and Back
TLDR
This paper attempts to identify from experiments the sources of instabilities in the controlled systems, and analyzes two types of policies: threshold-based and reinforcement learning techniques to dynamically scale resources.
Automated control of multiple virtualized resources
TLDR
Experimental evaluation with RUBiS and TPC-W benchmarks along with production-trace-driven workloads indicates that AutoControl can detect and mitigate CPU and disk I/O bottlenecks that occur over time and across multiple nodes by allocating each resource accordingly.
Auto-scaling to minimize cost and meet application deadlines in cloud workflows
  • Ming Mao, M. Humphrey
  • Computer Science
    2011 International Conference for High Performance Computing, Networking, Storage and Analysis (SC)
  • 2011
TLDR
This paper presents an approach whereby the basic computing elements are virtual machines (VMs) of various sizes/costs, jobs are specified as workflows, users specify performance requirements by assigning (soft) deadlines to jobs, and the goal is to ensure all jobs are finished within their deadlines at minimum financial cost.
...
1
2
...