Lecture Notes on Data Engineering

and Communications Technologies

Volume 162

Series Editor
Fatos Xhafa, Technical University of Catalonia, Barcelona, Spain
The aim of the book series is to present cutting edge engineering approaches to data
technologies and communications. It will publish latest advances on the engineering
task of building and deploying distributed, scalable and reliable data infrastructures
and communication systems.
The series will have a prominent applied focus on data technologies and commu-
nications with aim to promote the bridging from fundamental research on data
science and networking to data engineering and communications that lead to industry
products, business knowledge and standardisation.
Indexed by SCOPUS, INSPEC, EI Compendex.
All books published in the series are submitted for consideration in Web of Science.
Rajalakshmi Krishnamurthi · Adarsh Kumar ·
Sukhpal Singh Gill · Rajkumar Buyya
Editors

Serverless Computing:
Principles and Paradigms
Editors
Rajalakshmi Krishnamurthi Adarsh Kumar
Department of Computer Science School of Computer Science
and Engineering University of Petroleum and Energy Studies
Jaypee Institute of Information Technology Dehradun, Uttarakhand, India
Noida, India
Rajkumar Buyya
Sukhpal Singh Gill School of Computing and Information
School of Electronic Engineering Systems
and Computer Science The University of Melbourne
Queen Mary University of London Melbourne, VIC, Australia
London, UK

ISSN 2367-4512 ISSN 2367-4520 (electronic)

Lecture Notes on Data Engineering and Communications Technologies
ISBN 978-3-031-26632-4 ISBN 978-3-031-26633-1 (eBook)
https://doi.org/10.1007/978-3-031-26633-1

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2023
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

Serverless computing is a paradigm shift in cloud computing. Recently, many

companies rely on serverless computing for their product application develop-
ment and deployment, market analysis and customer relationship without investing
excess on infrastructure development and maintenance. This book brought a single
point of resource for researchers and practitioners on wide aspects of server-
less technologies. The book presents serverless computing, data-centric serverless
computing, distributed serverless computing and the road ahead and future of server-
less computing. Further, it focuses on the fundamental of serverless computing such
as the evolution of computing technologies, architecture, benefits, applications, issues
and solutions in serverless computing, open challenges and future scope. Further, the
book will present critical issues such as fine granularity and performance achieve-
ment in serverless computing. Next, the role of hyperscalers in serverless computing
in terms of application development, business and economic perspective will be
targeted.
The key performance concepts such as no operational costs, scheduling and
resource management, performance modelling, fairness, interoperability, virtual-
isation, data centres and portability are addressed. The merits of conventional
serverless computing include autoscaling and pay-as-you-go mode. It lacks effi-
cient data processing due to the shipping of data to the code, isolated VM for
serverless functions, and non-addressable and limited internal cache state. However,
modern computing in serverless requires data-intensive, distributed applications,
open-source platforms and customisable hardware. The topics under serverless data
lake architecture include functionalities such as data ingestion, data catalog, data
discovery/searching, ETL and ELT methodologies in serverless data lake archi-
tecture. Next, the containers orchestration on containers such as Docker, Kuber-
netes, and Linux Containers will be addressed. The commercial data-centric server-
less platforms frameworks such Amazon, Google, Azure and Oracle are covered.
This book also discusses the need for hardware-level enhancement for data-centric
serverless computing. For this purpose, the impact of multicore CPUs, cluster/grid
computing, graphic processing units, tensor processing units and FPGA accelera-
tors for serverless computing will be targeted. Further, the several big data format,

v
vi Preface

storage and services mechanisms for serverless computing are presented. Here, the
various modern data types and storage mechanisms such as spatial–temporal data,
time series data, key-value data, and graph-based data storage, columnar data storage,
real-time data streaming are addressed. The data-centric serverless services include
interactive queuing, real-time logging and monitoring, querying, and notification
services. Intensive data processing in serverless technology such as prediction, intel-
ligent decision-making, real-time, big data analytics, and data science support for
AI, ML and DL models in serverless computing is addressed.
This book focuses on distributed serverless computing. Here, the state manage-
ment, network file systems, communicating agents, autoscalability, P2P commu-
nication, generic- and application-specific frameworks, multi-tenancy and existing
distributed serverless computing frameworks are addressed. Further, the perfor-
mance issues in distributed serverless computing such as reliability, serviceability,
high availability, aggregation and broadcasting patterns, consistency, concurrency,
consensus, and fault-tolerant mechanism are addressed. Next, the data handling in
distributed serverless environments such as data sharing, replication, redundancy,
partitioning and indexing are addressed. This book addresses serverless technology
and primarily provides efficient mechanisms towards data privacy in terms of access
control auditing, attack and abuses. This book will also address the multiple serverless
computing and event-driven distributed systems. As a cutting-edge trend, serverless
computing is integrated with high-end computing technologies such as blockchain,
IoT, cloud computing, fog and edge computing, big data, artificial intelligence, SDN
and NFVs. This book serves as a platform for providing key insight and foreseen
open challenges towards serverless computing.
Chapters in this book are organised as follows:
The first chapter titled “Serverless Computing: New Trends and Research Direc-
tions” discussed that the serverless computing is an innovative method for the produc-
tion and distribution of software that does not rely on a centralised server manage-
ment infrastructure. Instead, the cloud service provider must ensure that the code
will execute as intended in the cloud environment. This frees up more time for the
developers to work on their projects. This chapter introduces serverless computing
and associated technologies. Additionally, this work provides future directions as
well as a summary of the research done for this book.
The second chapter titled “Punching Holes in the Cloud: Direct Communication
Between Serverless Functions” introduced a temporary architecture for function-
to-function communication in serverless systems through the use of direct network
connections. The framework has been successfully implemented on real, production-
ready serverless computing services, most notably AWS. To permit outgoing connec-
tions from functions while restricting inbound connections, contemporary serverless
computing systems frequently employ a networking configuration called network
address translation (NAT). Further, this chapter details the planning and development
of a library for transient communication in AWS Lambda. The network connection
between serverless applications is simplified by the library’s inclusion of function
and server components.
Preface vii

The third chapter titled “Hybrid Serverless Computing: Opportunities and Chal-
lenges” studied the extent to which serverless computing may be expanded to become
hybrid serverless computing. Further, the authors have defined hybrid serverless
computing, detailed the methods towards attaining it and highlighted the potential
and problems that it presents.
The fourth chapter titled “A Taxonomy of Performance Forecasting Systems
in the Serverless Cloud Computing Environments” focused on the classification
scheme used to characterise the parallel file system (PFS) structure. To understand
how existing PFSs are implemented in distributed computing environments and how
they might be adapted for usage in serverless (edge) cloud computing, a taxonomy
has been developed.
The fifth chapter titled “Open-Source Serverless for Edge Computing: A Tutorial”
investigated the options for deploying a serverless application at the edge of the
network using open-source software and Internet of things (IoT) data. Due to its focus
on resource economy and flexibility, the serverless method may be especially useful
for edge computing-based applications, in which the hosting nodes are deployed close
to the consumers and comprise devices and workstations with minimal resources.
The sixth chapter titled “Accelerating and Scaling Data Products with Server-
less” covered the framework and tools (data visualisation, pipelines, models, and
APIs) that help speed up and control data offers. APIs for data and model serving
with containerised solutions as a building block for data products that are driven by
machine learning techniques, and for serving a unified data ontology; data visualisa-
tion in the context of containerised web applications that deliver excellent methods
for data explorations, model predictions, visualisation and consumer insights.
The seventh chapter titled “QoS Analysis for Serverless Computing Using
Machine Learning” discussed the importance of artificial intelligence (AI) and
machine learning (ML) to make predictions regarding the system configurations
that are utilised in serverless computing. In addition, a model that does not incur
any costs is proposed to investigate and evaluate the many possible configurations
of workstations in an environment that lacks servers.
The eighth chapter titled “A Blockchain-Enabled Serverless Approach for IoT
Healthcare Applications” explored how blockchain technology might complement
serverless computing to address reliability issues with functions and resource alloca-
tion for IoT healthcare applications. The proposed method aims to react to customers’
demands in a trustworthy and dependable manner by taking their privacy concerns
into account, allocating resources efficiently, and meeting their needs promptly. It
is obvious that this paves the way for efficient use of resources, which in turn may
boost consumer happiness and service quality.
The ninth chapter titled “Cost Control and Efficiency Optimization in Main-
tainability Implementation of Wireless Sensor Networks Based on Serverless
Computing” provided a conceptual approach to the implementation of maintain-
ability for wireless sensor network (WSN) using serverless computing. To further
decouple the device operation and functional development, considerably optimise
viii Preface

the reuse of resources and remove the hardware interference, it has been proposed
that serverless computing may be accomplished at the software functional level
of WSN. To reduce design, manufacturing and operational costs, WSN platforms
may be built using the idea of serverless computing, which can support the func-
tions of data collecting and data management into functional development that may
benefit from exploration via upfront investments. Finally, a case study is provided
that uses existing technology and smart city scenarios to propose a WSN platform
for serverless computing.
The tenth chapter titled “Scheduling Mechanisms in Serverless Computing”
examined the benefits, drawbacks and uses of the most popular schedulers in server-
less computing. The current study’s goal is to give a thorough analysis of different
and efficient scheduling methods that can be used as a foundation for choosing the
right scheduling procedure based on the providers’ perspective.
The eleventh chapter titled “Serverless Cloud Computing: State of the Art
and Challenges” provided a thorough overview of these restrictions and to showcase
state-of-the-art research on ways to address the issues that are preventing server-
less from becoming the standard in cloud computing. The primary difficulties of
deploying such a cloud platform are examined, and potential research avenues are
outlined.
The book provides the best learning resource for researchers, graduates, under-
graduates, business people and common readers in the field of serverless computing.
When we talk about serverless cloud computing, it brings about tremendous changes
in the post-virtual-machine environment. Companies other than technology sectors
are using serverless platforms and frameworks at all production levels due to their
economic pay-per-use approach. Also, businesses of various shapes and sizes have
started to adopt serverless computing because of its scalability. Furthermore, the
technology’s use has enhanced IT infrastructures in the functions-as-a-service (FaaS)
sector. This enables a whole new range of workloads that are capable of benefiting
from the same capabilities of stateless programmes. It is now managed by a serverless
platform, so the burden of data management is removed for developers. This feature
helps business application development in a cloud-native way. This book acts as a
bridging information resource between basic concepts and advanced-level content
from technical experts to computing hobbyists towards enhancing their knowledge
and proficiency.

Noida, India Rajalakshmi Krishnamurthi

Dehradun, India Adarsh Kumar
London, UK Sukhpal Singh Gill
Melbourne, Australia Rajkumar Buyya
Contents

Serverless Computing: New Trends and Research Directions . . . . . . . . . . 1

Rajalakshmi Krishnamurthi, Adarsh Kumar, Sukhpal Singh Gill,
and Rajkumar Buyya
Punching Holes in the Cloud: Direct Communication Between
Serverless Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Daniel Moyer and Dimitrios S. Nikolopoulos
Hybrid Serverless Computing: Opportunities and Challenges . . . . . . . . . 43
Paul Castro, Vatche Isahagian, Vinod Muthusamy,
and Aleksander Slominski
A Taxonomy of Performance Forecasting Systems in the Serverless
Cloud Computing Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Sena Seneviratne, David C. Levy, and Liyanage C. De Silva
Open-Source Serverless for Edge Computing: A Tutorial . . . . . . . . . . . . . . 121
Priscilla Benedetti, Luca Gattobigio, Kris Steenhaut,
Mauro Femminella, Gianluca Reali, and An Braeken
Accelerating and Scaling Data Products with Serverless . . . . . . . . . . . . . . . 149
Angel Perez, Boyan Vasilev, Zeljko Agic, Christoffer Thrysøe,
Viktor Hargitai, Mads Dahlgaard, and Christian Røssel
QoS Analysis for Serverless Computing Using Machine Learning . . . . . . 175
Muhammed Golec, Sundas Iftikhar, Pratibha Prabhakaran,
Sukhpal Singh Gill, and Steve Uhlig
A Blockchain-Enabled Serverless Approach for IoT Healthcare
Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Mohsen Ghorbian and Mostafa Ghobaei-Arani

ix
x Contents

Cost Control and Efficiency Optimization in Maintainability

Implementation of Wireless Sensor Networks Based on Serverless
Computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Tinanan Gao and Minxian Xu
Scheduling Mechanisms in Serverless Computing . . . . . . . . . . . . . . . . . . . . 243
Mostafa Ghobaei-Arani and Mohsen Ghorbian
Serverless Cloud Computing: State of the Art and Challenges . . . . . . . . . 275
Vincent Lannurien, Laurent D’Orazio, Olivier Barais,
and Jalil Boukhobza