This book explores how advances in graphic processing units (GPUs), programmable logic devices (TPUs), and field-programmable gate arrays have altered the serverless computing landscape (FPGAs). Distributed system architectures and implementations have undergone significant changes due to the popularity of serverless computing. Making and releasing product applications, doing market research, and maintaining customer interactions might all benefit from the reduced infrastructure expenses made possible by serverless computing. This book is a great resource for teachers and students interested in learning more about serverless computing. Some of the main questions surrounding serverless technology, such as scalability and performance distribution, are answered. Concepts and fundamentals of computing performance such as cost-free operation, good time and resource management, fairness, and interoperability are discussed. Serverless is at the forefront of this shift, which has made data-intensive, distributed applications, and open-source platforms essential for any modern computer to function. Data-centric queuing, real-time logging and monitoring, querying, and alarms are all examples of serverless services.
Author(s): Rajalakshmi Krishnamurthi, Adarsh Kumar, Sukhpal Singh Gill, Rajkumar Buyya
Series: Lecture Notes on Data Engineering and Communications Technologies, 162
Publisher: Springer
Year: 2023
Language: English
Pages: 319
City: Cham
Preface
Contents
Serverless Computing: New Trends and Research Directions
1 Introduction
1.1 Motivation
1.2 Traditional Versus Serverless Computing
2 Serverless Functions, Architectures and Computing
3 Resource Management in Serverless Computing Environments
4 Serverless Computing and Advanced Technologies
5 Open Challenges of Resource Management in Serverless Computing
6 Future Directions
7 Conclusion
References
Punching Holes in the Cloud: Direct Communication Between Serverless Functions
1 Introduction
1.1 Novelty
1.2 Design and Contributions
2 Background
2.1 Serverless Computing
2.2 Data Storage with Serverless
2.3 Network Address Translation (NAT)
2.4 Motivation for This Work
3 Design
3.1 Communication Library Functionality
3.2 Client-Server Design
3.3 NAT Traversal Protocol
4 Implementation
4.1 NAT Traversal Technique
4.2 Server-Function Communication
4.3 Communication Library API Overview
4.4 Function Program Design
5 Evaluation
5.1 TCP Performance in Lambda
5.2 TCP Versus S3 Comparison
6 Related Work
6.1 Serverless Communication
6.2 Serverless Execution Frameworks
6.3 Serverless Applications
6.4 FaaS Platforms
7 Conclusion
7.1 Limitations and Future Work
7.2 Summary
References
Hybrid Serverless Computing: Opportunities and Challenges
1 Introduction
2 Trends in Cloud and Serverless Computing
2.1 Cloud and Hybrid Cloud
2.2 Computing Trends Toward Hybrid Computing
3 The Case for Hybrid Serverless Computing
3.1 Hybrid Serverless Architecture Design Principles
3.2 Benefits to Application Developers and Platform Providers
3.3 Roadblocks to Hybrid Serverless Computing
4 Towards a Hybrid Serverless Architecture
4.1 Two Approaches to Achieving Hybrid Serverless Computing: Compilers and Standards
4.2 Discussion
4.3 Vision
5 Opportunities and Challenges
5.1 Standards
5.2 Foundational Challenges in Hybrid Serverless Computing
5.3 High-Impact Challenges in Academia and Industry
6 Conclusion
References
A Taxonomy of Performance Forecasting Systems in the Serverless Cloud Computing Environments
1 Introduction
2 Background and Related Work
3 Challenges
4 The Taxonomy
4.1 Classification of PFS by the Design Methodology of Forecasting Algorithms
4.2 Forecasting Approaches Trained from Historical Execution
4.3 Resource Type Taxonomy (Resource Type Hierarchy Tree)
5 Survey
5.1 The Surveyed Performance Forecasting Systems
6 Analysis of the Survey
6.1 Introduction
6.2 Meeting the Challenges
6.3 Taxonomy
7 Conclusions
References
Open-Source Serverless for Edge Computing: A Tutorial
1 Introduction
2 Open-Source Serverless Platforms
2.1 A Taxonomy of Open-Source Serverless Platforms
2.2 Deploying a Serverless Application with OpenFaaS
3 Beyond Containers: Lightweight Virtualization Techniques
3.1 Unikernels
3.2 Kata Containers
3.3 Integration in Serverless Platforms
4 Serverless Security for Edge Computing
4.1 Serverless Security Advantages for Edge Computing
4.2 The Tradeoff Between Security and Performance
4.3 The Cold Versus Warm Start Problem
5 A Focus on IoT Scenarios
6 Handling Cloud-Edge Interworking
6.1 Connecting Cloud and Edge
6.2 Cloud and Edge Interactions
7 Conclusion
References
Accelerating and Scaling Data Products with Serverless
1 Introduction
2 Data Products and Their Architecture
3 Serverless Building Blocks for Data Products
3.1 Data Ingestion, Storage, and Transformation
3.2 Models and Data Serving Using APIs
3.3 Data Visualization and User Interfaces
4 Real-World Examples, Data Products Catalogue
4.1 Data Processing Foundation for Analytics
4.2 Ontology Visualization and Business Logic Compliance
4.3 Community Analytics
4.4 Recommendation System
4.5 Data Lineage Visualization
4.6 Discussions
5 Summary and Concluding Remarks
References
QoS Analysis for Serverless Computing Using Machine Learning
1 Introduction
1.1 The Trend in CPU and GPU Performance
1.2 Motivation
2 Background
2.1 Beginning of Serverless
2.2 CPU and GPU
3 Theory
3.1 Moore's Law
3.2 Regression Model
3.3 R2 Score
4 Methodology
4.1 Dataset
4.2 Preprocessing and Feature Engineering
4.3 Calculation and Fitting of Curve
5 Evaluation and Results
5.1 Theoretical Moore's Law
5.2 Fitted Moore's Law
5.3 Two-Degree Polynomial
5.4 Three-Degree Polynomial
5.5 Graphical Representation
5.6 Experimental Results
6 Discussion
6.1 COVID-19 and It's Impact
6.2 Drawbacks
7 Conclusions
8 Future Work
References
A Blockchain-Enabled Serverless Approach for IoT Healthcare Applications
1 Introduction
2 Preliminaries
2.1 Serverless Computing
2.2 Scheduling in Serverless Computing
2.3 Blockchain
3 Blockchain for Scheduling in Serverless Computing in the Healthcare Domain
3.1 Architecture Hyperledger Fabric a Permission Blockchain
3.2 Architecture Serverless Computing
3.3 Interacts Serverless Computing with Hyperledger Fabric Blockchain
4 Combining Blockchain and Cloud Computing Environments
5 Scheduling in Serverless Computing with Hyperledger Fabric
6 Conclusion
References
Cost Control and Efficiency Optimization in Maintainability Implementation of Wireless Sensor Networks Based on Serverless Computing
1 Introduction
2 WSN Maintainability Implementation
2.1 Definition
2.2 Life Cycle Phase
2.3 Resources and Costs
2.4 Processes and Features
2.5 Current Situation and Challenges
3 Serverless Computing
3.1 Concept and Application
3.2 Serverless Computing Used in WSN's Maintainability
4 Implementation Conception
4.1 Existing Schemes and Problems
4.2 Resource Provisioning Strategy
5 Conclusions and Future Directions
References
Scheduling Mechanisms in Serverless Computing
1 Introduction
2 Preliminaries
3 Scheduling in Serverless Computing
3.1 Scheduling Architecture in Serverless Computing
4 Scheduling Algorithms in Serverless Computing
4.1 Energy-Aware Scheduling in Serverless Computing
4.2 Data-Aware Scheduling in Serverless Computing
4.3 Deadline-Aware Scheduling in Serverless Computing
4.4 Resource-Aware Scheduling in Serverless Computing
4.5 Package-Aware Scheduling in Serverless Computing
4.6 Cost-Aware Scheduling in Serverless Computing
4.7 History-Aware Scheduling in Serverless Computing
5 Open Issues and Future Directions
6 Discussion and Comparison
6.1 Programming Language
6.2 Cloud Execution Layers
6.3 Frameworks
6.4 Performance Metrics
7 Conclusion
References
Serverless Cloud Computing: State of the Art and Challenges
1 Introduction
2 Background and Motivations
2.1 The Promises of Cloud Computing
2.2 Virtualization Technologies
2.3 From Monoliths to Microservices
3 Serverless, A New Paradigm
3.1 Characteristics of Serverless Platforms
3.2 Suitable Workloads
3.3 Tradeoffs in Serverless Deployments
3.4 Description of Current FaaS Offerings
4 Problems Addressed in the Literature
4.1 Cold Start Delays and Frequency
4.2 Data Communications Overhead
4.3 Durable State and Statefulness
4.4 Hardware Heterogeneity
4.5 Isolation and Security
4.6 Programming Model and Vendor Lock-in
5 Perspectives and Future Directions
5.1 Service Level Agreements
5.2 Energy Efficiency
5.3 AI-Assisted Allocation of Resources
6 Conclusion
References
