Hassan et al.

Journal of Cloud Computing: Advances, Systems Journal of Cloud Computing:

and Applications (2021) 10:39
https://doi.org/10.1186/s13677-021-00253-7 Advances, Systems and Applications

REVIEW Open Access

Survey on serverless computing

Hassan B. Hassan1 , Saman A. Barakat2 and Qusay I. Sarhan2*

Abstract
Serverless computing has gained importance over the last decade as an exciting new field, owing to its large
influence in reducing costs, decreasing latency, improving scalability, and eliminating server-side management, to
name a few. However, to date there is a lack of in-depth survey that would help developers and researchers better
understand the significance of serverless computing in different contexts. Thus, it is essential to present research
evidence that has been published in this area. In this systematic survey, 275 research papers that examined serverless
computing from well-known literature databases were extensively reviewed to extract useful data. Then, the obtained
data were analyzed to answer several research questions regarding state-of-the-art contributions of serverless
computing, its concepts, its platforms, its usage, etc. We moreover discuss the challenges that serverless computing
faces nowadays and how future research could enable its implementation and usage.
Keywords: Survey, Cloud computing, Serverless computing, Serverless platforms, Serverless benefits, Serverless
challenges

Introduction Meanwhile, in the PaaS, cloud companies provide services

Cloud computing emerged after the appearance of virtu- such as network access, storage, servers, and operating
alization in software and hardware infrastructures; hence systems to be purchased by developers. The developers
cloud providers increasingly adopted it to offer their ser- access these services to deploy, run, and manage their
vices to customers [1, 2]. Customers can access these applications. In this kind of cloud, the developer is respon-
cloud services via the Internet. Software developers have sible for the deployment and management (settings and
been using cloud technologies in their software solutions configurations) of their software to ensure that the appli-
owing to their benefits including scalability, availability, cation is running, while they do not control the services.
and flexibility [3]. Finally, in the IaaS category, the cloud consumers con-
In general, cloud computing is divided into three main trol and manage services such as network access, servers,
categories based on the provision of services, which are operating systems, and storage.
software as a service (SaaS), platform as a service (PaaS), Managing cloud services is not an easy task at all. The
and infrastructure as a service (IaaS). In the SaaS cate- authors in [4] have addressed several challenges while
gory, cloud providers offer different types of software as managing a cloud environment by a user such as availabil-
services to the users. For example, Google provides many ity, load balancing, auto-scaling, security, monitoring, etc.
applications as a service (e.g., Gmail, Google docs, Google For example, the cloud user has to ensure the availability
sheets, and Google forms). In this type of cloud, the user of the services in which if a single machine failure occurs,
is not responsible for the services development, deploy- it does not affect the whole services. Also, he/she has to
ment, and management. The user here only uses them consider distributing copies of the services geographically
without worrying about their settings, configurations, etc. to protect them when disasters happen. Another challenge
is load balancing. In this case, the cloud user has to ensure
*Correspondence: [email protected] that requests to the services are balanced to utilize all
2
Software Engineering and Embedded Systems (SEES) Research Group,
Department of Computer Science, College of Science, University of Duhok, resources efficiently.
Duhok, Kurdistan Region, Iraq
Full list of author information is available at the end of the article

© The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit
to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The
images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 2 of 29

These challenges have led to introduce another cloud to concern themselves about the scaling issues. For exam-
computing model, which is called serverless cloud com- ple, when an application runs on a serverless cloud, it
puting [4]. Serverless cloud computing offers backend as a will scale up automatically when the application requests
service (BaaS) and function as a service (FaaS), as shown increase. Another characteristic of serverless computing
in Fig. 1. The BaaS includes services like storage, mes- is the payment per resource usage. This paradigm of
saging, user management, etc. While, the FaaS enables cloud computing charges developers based on the actual
developers to deploy and execute their code on comput- resource usage. For example, deploying an application will
ing platforms. The FaaS relies on the services provided by not cost the developer in the case where the application is
the BaaS such as a database, messaging, user authentica- idle, and the serverless provider will only charge whenever
tions, etc. The FaaS is considered as the most dominant the application has started using resources.
model of serverless, and it is also known as “event-driven However, any new technology will face numerous tech-
functions” [5, 6]. nical and operational issues and obstacles at the begin-
Serverless cloud model was for the first time introduced ning. Since the recent introduction of serverless cloud
by Amazon Lambda in 2014, after which cloud companies computing, several drawbacks have been identified [7].
like Google and Microsoft adopted it in 2016. Serverless Serverless cloud computing lacks tools that help manag-
cloud computing adds an additional abstraction layer to ing and monitoring serverless applications. Moreover, it
the existing cloud computing paradigms, while it abstracts might comprise security concerns. Further, the serverless
away the server-side management from the developers [7]. providers have a vendor lock-in problem. Nevertheless,
Serverless model lets the developers focus on the appli- serverless cloud computing has gained positive atten-
cation logic rather than the server-side management and tion in the industry, despite that it has not been studied
configurations. For example, the developers deploy their extensively in academic research [7].
applications to the serverless cloud as functions see Fig. 1. Therefore, the aim of this research is to answer some
Then, the cloud provider takes responsibility for manag- crucial research questions related to serverless cloud com-
ing, scaling, and providing different resources to ensure puting and thereby help researchers as well as developers
the smooth running of these functions [8, 9]. to better understand serverless cloud computing and con-
However, FaaS and the term “serverless” could be used tribute to its development.
interchangeably, as the FaaS platform automatically con- The rest of this paper is structured as follows: “Related
figures and maintains the execution context of functions works” section presents the related works for this study.
and connects them to cloud services without requiring “Research methodology” section describes in detail the
server provision by developers [10, 11]. We refer to the research methodology used to conduct this survey study.
FaaS when we use the term serverless computing. “Results” section presents the results and outcomes of the
Serverless cloud computing has many good character- study. “Threats to validity” section presents the threats to
istics [12, 13], one of which is scalability. Scaling could validity of this study. Finally, the conclusions of the study
be vertical or horizontal; vertical scaling adds or removes are provided in “Conclusions” section.
cores from the running container, while horizontal scal-
ing creates new containers or eliminates running ones Related works
without affecting the current resource allocations [14]. In The most relevant studies published on the topic are
serverless computing, the applications automatically scale briefly presented here. The authors in [15] and [16] dis-
up and down on demand, and the developer does not have cussed some important background to the origin and

Fig. 1 Serverless architecture

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 3 of 29

evolution of serverless computing and the long road bibliometric, content, technology, and produced statistics
that serverless computing has taken over the years. The about each section. In contrast, our paper aims to conduct
authors in [9] thoroughly discussed the true meaning of a systematic survey. In this survey, we try to find answers
serverless architectures and how they are changing the to several critical questions related to serverless comput-
way in which applications are built, deployed, and dis- ing. In addition to that, our study covered the duration
tributed. (2016–2020) and thus 275 papers have been considered.
Numerous studies focused on technical interpretations The authors in [25] mainly focused on scheduling tasks
of serverless computing, while other more recent research in the cloud. They described the various techniques in
suggested various benefits that it brings to developers. scheduling workflows to reduce the execution time, cost,
Nowadays, this type of computing is being used in several or both. Moreover, they proposed a hybrid method by
ways. In an empirical study, the authors in [17] aimed to both FaaS and IaaS. The small tasks could be executed
investigate the development practices of serverless com- remotely using the FaaS, which reduces the execution
puting in the industry. They concluded that for develop- cost; hence, the number of virtual machines will be
ers, it remains a barrier to adopt the right mindset to best decreased as well. Therefore, the whole focus would be on
utilize the tools inherent to serverless architecture. More the long-running tasks on IaaS.
documentation and easier access to such resources would The authors in [26] covered only 24 research papers
help developers to embrace the possibilities that serverless during 2017–2019. In their paper, they considered both
computing has to offer. the white and grey literatures. Besides, they identified 32
The concept of serverless computing within the scope characteristics of serverless and the possible issues related
of the IT industry has the great potential of progres- to them, only eight of them were stated by both white
sively increasing its capabilities to involve a wider set of and grey literatures while the remaining are from grey
domains. Thus, the implementation of serverless comput- literature only. All the characteristics are explained and
ing is not restricted only to the enhancement of infras- presented briefly. In our paper, 275 research papers from
tructure, and it can be employed for many different pur- 2016–2020 have been covered and more research ques-
poses, e.g., serverless messaging, neural network training tions have been answered. Besides, a well-defined system-
[18], video processing [19], and big data [20]. Undeniably, atic literature study process has been employed. Thus, the
their contributions are valuable to the general public and grey literature has been excluded in our paper and, our
researchers in the field, as it is of primarily importance to results are reproducible compared to their results.
comprehend how this technology works. The authors in [27] mainly concentrated on difficul-
However, it is presently crucial to provide more than ties and gaps in data-centric and distributed computing
only theories and concepts: it is time to weigh the bene- using FaaS. Additionally, they evaluated the severity of
fits and drawbacks of serverless computing and to analyze these challenges via taking three case studies from big data
how far the field has progressed, to assess what remains to and distributed computing settings: model training, low-
be done and improved. As an example, the authors in [21] latency prediction serving using the batch and, distributed
discussed some possible new abstraction levels to reduce computing. While our paper is a broad and comprehen-
processing limitations. The authors in [22] discussed the sive study on FaaS, 275 research papers are taken from the
results from an open-source framework to achieve on- white literature during 2016–2020.
premises serverless computing that can process big work- The paper [28] presented only four use cases of FaaS:
loads with a scalable and sensible usage of resources. We event-triggered computing, video broadcasting, Internet
can infer from these related publications that researchers of Things (IoT) data processing, and shared delivery sys-
everywhere are working to determine how to best exploit tem. Additionally, the paper only compared three plat-
the potential that serverless computing frameworks could forms namely, Amazon web services (AWS) Lambda,
introduce to software development. Google Cloud Function, and Microsoft Azure Func-
In [23], the authors described how serverless computing tion. On the other hand, our paper presents a com-
is becoming the next step in the evolution of cloud com- prehensive study about FaaS. We identified in detail
puting and its platforms. In our paper, we focus on the eight use cases: chatbot, information retrieval, file pro-
ongoing challenges, benefits, and drawbacks of using it. cessing, smart grid, security, networks and, mobile and
The authors in [24] have conducted a systematic explo- IoT. Moreover, our paper compared ten FaaS platforms
ration of serverless computing-related research papers. As namely, AWS Lambda, Apache OpenWhisk, Microsoft
they mentioned, their work is not a survey, but it is a Azure functions, Google Cloud functions, OpenLambda,
supporting source for future research papers. They pro- IBM Cloud functions, OpenFaaS, Knative, FunctionStage,
posed an open dataset for serverless computing papers. Huawei Cloud, and Nuclio.
The dataset includes 60 papers for the period (2016-July The authors in [29] covered only 15 papers during
2018). Also, they have analyzed the dataset according to 2016–2018. They took both the white and grey literatures

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 4 of 29

into account. On the other hand, our paper includes paper has covered 275 research papers from 2016–2020
275 research papers published in the period 2016–2020; forming a well-defined systematic literature study. We also
they are taken from the white literature only. Moreover, identified 21 serverless challenges and issues. Besides, we
our paper has formulated and answered eight clear and compared serverless with the traditional cloud computing
well-defined research questions. paradigm. We identified more application areas including,
The authors in [30] focused on the FaaS performance chatbot, information retrieval, file processing, smart grid,
evaluation and their publication trends during 2016– security, networks, IoT, and edge computing.
2019. They identified the most commonly evaluated FaaS The authors in [31] presented a white paper based on
platforms. Additionally, they evaluated the performance published research papers during 2015–2017. They out-
features for benchmark types, micro-benchmarks, and lined the serverless definition alongside its advantages
common features across FaaS platforms. Moreover, they and disadvantages. Also, they classified serverless use-
presented the most common platform configurations in cases into six categories, namely, backends, web appli-
FaaS, namely language runtimes, function triggers, and cations, chatbots, big data, IT automation, and Amazon
external services. This paper presents a survey of the most Alexa. Moreover, they addressed a few research ques-
important and state of the art aspects of FaaS. Besides, tions including, the need for the stateless feature in
comprehensive theoretical aspects of FaaS are covered serverless, whether serverless could execute long-running
taking from the white literature during 2016–2020. tasks, programming models, serverless standards, and the
The authors in [11] have conducted a systematic map- importance of serverless for scientific research. While
ping study on serverless cloud computing. The main aim our paper is a comprehensive study on FaaS; we cov-
of their study is to concentrate on FaaS engineering. ered 275 research papers which are taken from the grey
They raised two main concerns: (a) identifying publi- literature during 2016–2020. In our paper, eight applica-
cation research that considers developing or modifying tion areas have been identified as mentioned earlier. We
serverless platforms and tools. (b) identifying the chal- have identified and answered ten research questions that
lenges and drivers related to these publications. On the cover many aspects of the topic in detail compared to the
other hand, our study extends the challenges and issues aforementioned study.
related to serverless computing. Moreover, we provide We are in fact addressing with this paper ten important
more details about serverless computing platforms and research questions about the topic, potentially making it
the use of these platforms in the literature. Also, it pro- a more complete guide to the development and use of
vides a detailed comparison among the most widely used serverless computing. Our work contributes to the anal-
serverless platforms. Besides, it addresses more aspects ysis of the serverless paradigm in the context of similar
of serverless cloud computing such as application areas applications and how could they better fit specific com-
of serverless computing, future directions of serverless puting needs. Moreover, information about the current
computing, etc. state of serverless platforms, tools, and frameworks has
The authors in [7] provided useful observations about been updated for this survey. This due to the importance
serverless computing, its architecture, and use cases. Also, of the topic and its potential to change how both the
they discussed the challenges and benefits of moving for- industry and academia have managed the deployment of
ward from monolithic applications and the differences cloud applications until now. Updated information about
between traditional cloud services and serverless com- the area could benefit future studies focused on the server-
puting. Our work has extended the details of their work less computing paradigm as they make researchers aware
regarding the benefits and drawbacks of using server- of the latest resources and opportunities in the area.
less computing. It has also included more use cases and
workloads to deepen the findings of previous studies. Research methodology
The authors in [4] presented a technical report on Research questions
serverless computing. They covered the serverless emer- In this study, a number of research questions (RQs) have
gence with its limitations, including limited storage for been identified and answered. Each RQ addresses a par-
fine-grained tasks, lack of coordination among func- ticular aspect of serverless computing as follows.
tions, inadequate performance for standard communi-
cation patterns, and functions’ performance. Also, they • RQ1. What is the number and distribution of studies
compared AWS serverful with AWS serverless. More- published on serverless computing in the period
over, they also explained the challenges of architecture, (2016–2020)?
networking, security, and abstractions of serverless com- • RQ2. Which researchers, organizations, and
puting. They identified five application areas including, countries are active in serverless computing research?
video encoding in real-time, MapReduce, linear alge- • RQ3. What are the differences between serverless
bra, machine learning training, and databases. While our computing and traditional cloud computing?

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 5 of 29

• RQ4. What are the benefits of using serverless inaccurate results have been obtained due to the partial
computing? similarity to FaaS, such as the federal aviation adminis-
• RQ5. What are the most used software platforms that tration (FAA). The results of the initial search were 5,021
enable serverless computing in the literature? papers in total.
• RQ6. What are the application areas of serverless After obtaining the initial list of publications, some fil-
computing in the literature? ters have been applied to reduce the number of incorrect
• RQ7. What are the challenges and issues of using results based on their relation to the serverless computing
serverless computing? and FaaS topics. Most of the papers have been analyzed
• RQ8. What tools are available for serverless based on the title and abstract. However, when we were
computing? (serverless tools) unable to make a decision based on the title and the
• RQ9. What are the available research approaches to abstract, we read the content of the paper to ascertain
analyze the migration of monolithic applications to whether to include or exclude. As a result, the list of
serverless computing? papers which are related to serverless computing has been
• RQ10. What are the potential future directions of decreased to 549 papers.
research on serverless computing? After filtering the papers based on the title and abstract,
we merged all the papers that were relevant to server-
Search strategy less cloud computing, which was 549 papers into a single
Literature sources dataset. Then we removed the duplicated papers based on
In this study, five standard online databases have been the combination of a title, author names, publication year,
selected as sources that index the literature of software and venue. Thus, the number of publications has been
engineering and computer science. These sources are pre- reduced to 489 papers.
sented in Table 1. Then, the publications have been selected based on
the content of the paper and based on a set of inclu-
Search string sion/exclusion criteria (see the following section) that
To find the publications relevant to this study, the fol- have been selected carefully. Eventually, we could obtain
lowing extensive search string has been applied on the 254 papers that are related to serverless cloud comput-
database sources of literature: ing. In the next step, we applied backward snowballing to
(serverless OR FaaS OR “function as a service” OR increase the set of relevant papers to serverless cloud com-
“function-as-a-service”) AND (computing OR paradigm puting. In this phase, we could add 21 more papers to our
OR architecture OR model OR application OR function OR list of papers. As a result, the total numbers of relevant
service OR platform OR programming) papers become 275 papers. The list of these papers and
To obtain the best publication list, a generic search its meta-data have been published in Zenodo website as a
string is created. It contains serverless cloud computing- dataset [32].
related keywords. The string with duration (2016 - 2020)
have been applied to all libraries. Because the Springer
Paper inclusion/exclusion criteria
Link library covers many fields, the result of search was
To decide whether a publication is relevant to the scope of
greater than other libraries. This because the keyword
this research, a set of inclusion and exclusion criteria have
FaaS is used in many research areas for different pur-
been established and employed as follows:
poses. For instance, fish as a service (FaaS) and FPGA as a
fervice (FaaS). Therefore, we used Computer Science sub- Inclusion criteria:
ject filter with Springer Link, ScienceDirect, and Scopus • Publications in the field of software engineering and
to reduce the number of incorrect papers. The results of
computer science.
the initial search are shown in Fig. 2. Additionally, some • Publications published online from 2016 – 2020.
• Publications related directly to serverless computing.

Exclusion criteria:
Table 1 Database sources used to explore the literature
Source URL • Publications not published in English.
IEEE Xplore http://ieeexplore.ieee.org
• Publications without accessible full text.
• Publications not formally peer reviewed (e.g., gray
Elsevier ScienceDirect http://sciencedirect.com
literature).
ACM Digital Library http://portal.acm.org • Publications not published electronically.
Scopus http://scopus.com • Publications that are duplicates of other previous
SpringerLink http://springerlink.com publications.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 6 of 29

Fig. 2 Results of papers selection process

Results the billing is only on the basis of function execution

The selected publications were carefully read to answer time and resource allocation. Also, the developers are
the raised RQs. Here, a short title is used to represent not required to be aware of the underlying infrastruc-
each RQ. The following subsections present and discuss ture and workflows. Hence, this attracts cloud providers
the results based on each RQ. and businesses to migrate and support serverless along-
side many directions. At the same time, researchers are
Distribution of publications (RQ1) paying more attention to serverless as it is becoming the
Publication frequency future paradigm for cloud computing. Moreover, current
All the selected papers of this study were analyzed to challenges and limits in serverless computing draw atten-
determine their frequency and evolution. Figure 3 shows tion to more academics to address the issues and enhance
the results of this analysis. The results show that the aver- the currently available features. For the aforementioned
age number of publications per year is approximately 55 reasons, developers and customers are well encouraged
papers. and satisfied to select serverless computing for developing
Serverless computing has trended a significant engage- applications and services.
ment over the past two years. This boost has been caused
by industry, academia, and developers for several rea- Publication venue
sons. The first important reason is the attractive engage- The distribution of the selected papers in various pub-
ment opportunities that serverless offers cloud providers. lication venues is shown in Fig. 4. The percentages of
Serverless nature equipped cloud providers with more publications in conference papers, workshop papers, sym-
convenient and efficient methods to manage and uti- posium papers, and journal papers are approximately 62%,
lize idle computing resources. Another reason is that 11%, 14%, and 13%, respectively. However, almost 13% of

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 7 of 29

Fig. 3 Published papers per year

the studies have been published in journals, which indi- “Cloud”, “UCC”, “SoCC”, and “Middleware” are consid-
cates the immaturity of research in serverless computing ered the most active conferences that hold approx-
[33, 34]. It is worth mentioning that some conference imately 28% of the published conference papers. By
papers were published as book chapters. Thus, the orig- including other conferences with three published papers
inal venues, which are conferences, of such papers were or more, then approximately 23% of the conference
considered. papers are published in annual conferences. The major-
Following the interpretation of publications, the most ity (almost 49%) of the conference papers were published
productive and primary journals, symposiums, confer- at individual conferences, which are denoted as “Others”
ences, and workshops venues related to serverless com- in Fig. 6.
puting can be clarified. Due to their long names, abbre-
viations are used in this paper. The active journals are Active researchers (RQ2)
shown in Fig. 5 and their full names can be found in Serverless computing is a vital research area through
Table 2. It can be observed from the figure that the top the contribution of several scholars. Yet, the researchers
and vital three journals are “FGCS”, “IoT”, and “JSS”. Also, are counted active if they contributed to more than
it can be noticed that the top three journals contain almost two research studies, as presented in Fig. 7. The figure
34% of the published journal papers, while the others own shows that the top six active researchers are “Pedro Gar-
approximately 66%. cáa López”, “Erwin Van Eyk”, “Alexandru Iosup”, “Marc
The active conferences are shown in Fig. 6 and their Sánchez-Artigas”, “Sebastian Werner”, and “Wes Lloyd”.
full names are presented in Table 3. The “WOSC”, Table 4 presents the active nations, research institutions,

Fig. 4 Published papers ratio per each venue

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 8 of 29

Fig. 5 Published papers vs. journal name

researchers, references to the published papers, and the logic into smaller functions increase the development
total number of publications. efficiency [77, 177] and also decreases the chance of a
The active nations in the number of papers are obtained single point of failure [77]. On the other hand, the com-
from the information presented in Table 4 by extracting ponent dependency within monolithic applications affects
the institutional affiliation of the authors and co-authors. the availability of other services adversely.
An overview of the most active nations and the total num- In a serverless architecture, the developers are unable to
ber of publications is shown in Fig. 8. It is observable that take control of listening to the TCP socket, managing load
the United States and Germany are the largest contribu- balancers, maintenance or configuration of the server, as
tors to papers published on serverless computing with 104 well as provisioning and resource allocation. Therefore,
and 39 published papers, respectively. there is no need for system administrators; the devel-
opers only focus on handling client requests and paying
Serverless computing vs. traditional cloud attention to deliver valuable services [8].
computing (RQ3) There are several differences between Serverless computing also differs from monolithic com-
serverless and traditional cloud computing. In the tra- puting as the functions have shorter life cycles.
ditional cloud architecture, the server acts as a mono- The traditional monitoring and debugging tools that are
lithic system containing all business logic. Meanwhile, used in monolithic applications are not included in the
the serverless architecture is modeled into smaller, event- serverless architecture; the developers are compelled to
driven, and stateless ‘triggers’ (events) and ‘actions’ (func- use built-in tools for debugging and monitoring. The com-
tions) [175]. Each component handles different pieces of puting power is no longer a concern for the developers
data and runs independently [176]. Spreading business in the serverless paradigm, as it could scale horizontally
almost indefinitely [178, 179]. Meanwhile, in the client-
server architecture, it usually requires dedicating two
Table 2 List of active journals
server instances; the primary instance and a second in
Acronym Journal Full Name case if the former fails. This leads to higher costs in the
FGCS Future Generation Computer Systems monolith paradigm. Serverless architecture could be more
IoT-J Internet of Things Journal economical for unsteady load conditions while the server-
JSS Journal of Systems and Software based is more suitable for steady loads [152]. As serverless
applications scale up and down according to the requests,
Clust. Comput. Cluster Computing
thus, unlike the traditional systems, it is unnecessary to
IEEE Access IEEE Access
keep the sessions in the memory [8]. Hence, it is difficult
IEEE Internet Comput. IEEE Internet Computing to keep track across requests.
PACMPL Proceedings of the ACM on Programming FaaS boosts the security level as cloud providers contin-
Languages uously update their infrastructure with the latest security
SICS Software-Intensive Cyber-Physical Systems patches; this also removes the security burden on devel-
TPDS IEEE Transactions on Parallel and Distributed opers [17]. Directly accessing the backend resources in the
Systems traditional model is considered a critical security issue.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 9 of 29

Fig. 6 Published papers vs. conference name

Thus, any requests from the clients and internal func-

Table 3 List of active conferences tions in the serverless environment must go through a
distributed request-level authorization mechanism that
Acronym Conference Full Name
strengthens the security level [8]. Additionally, denial of
WOSC International Workshop on Serverless Computing
(WOSC)
service (DoS) attacks are controlled, as it is more diffi-
cult to attack distributed servers than a single server [175].
Cloud International Conference on Cloud Computing
(Cloud) However, some security concerns remain due to the third-
UCC International Conference on Utility and Cloud party API usage [9]. Besides, there is a lack of tools to
Computing (UCC) identify vulnerabilities and access control risks. Table 5
SoCC Symposium on Cloud Computing (SoCC) summarizes the aforementioned differences.
Middleware International Middleware Conference (Middleware)
Benefits of serverless computing (RQ4)
IC2E International Conference on Cloud Engineering (IC2E)
Serverless computing offers numerous benefits to its
ICPE International Conference on Performance
Engineering (ICPE) users, and Table 6 presents papers that states these bene-
USENIX ATC USENIX Annual Technical Conference (USENIX ATC) fits. This section summarizes those benefits as follows:
CCGRID International Symposium on Cluster, Cloud and Grid Cost effective
Computing (CCGRID)
Serverless applications are abstracted from server infras-
DEBS International Conference on Distributed and
Event-based Systems (DEBS) tructure, which results in cost-based services depending
ESOCC European Conference on Service-Oriented and Cloud
on usage [180]. For example, applications run whenever a
Computing (ESOCC) user makes a request to a service within the application.
HotCloud Conference on Hot Topics in Cloud Computing The cloud vendors charge only for the used space, and
(HotCloud) there is no cost while their applications are in an idle state.
ICDCS International Conference on Distributed Computing
Systems (ICDCS) Scalability
ICDCSW International Conference on Distributed Computing Serverless reasonably solved the resource allocation prob-
Systems Workshops (ICDCSW) lem [191]. Therefore, developers do not have to con-
ICFC International Conference on Fog Computing (ICFC) cern themselves with the application scalability, because
JCC International Conference on Joint Cloud Computing the application will scale automatically whenever user
(JCC) application requests are increased. If there are numerous
SAC Symposium on Applied Computing (SAC) requests to a function within the application, the server-
SEC Symposium on Edge Computing (SEC) less providers will start servers to handle these requests.
SYSTOR International Systems and Storage Conference
(SYSTOR) Server-side management
XP International Conference on Agile Software In serverless computing, developers do not need to con-
Development (XP) cern themselves with the server-side and its management.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 10 of 29

Fig. 7 Active researchers based on the published papers

Serverless cloud providers take care of managing and user requests do not have to travel across the Internet to
maintaining the hardware and software required to deploy access the original server.
applications. In addition to that, they handle all admin-
istration operations to let developers focus on different Serverless platforms in the literature (RQ5)
kinds of resources such as central processing unit (CPU), The software platforms are generally implemented to deal
memory, and storage. with resources from several clouds and ensure proper run-
ning of client applications. The heterogeneous nature of
Easy to deploy the cloud providers’ infrastructure (hardware and operat-
Serverless applications are easy to deploy. For example, ing systems) led to the necessity to direct the developers’
to deploy an application, developers only need to upload focus to the functional part, rather than the underlying
some functions and release a new product. The serveless infrastructure [199].
will take care of deployment management and infras- With the emergence of the first serverless platform,
tructure related concerns such as server provisioning and AWS Lambda by Amazon in 2014 [8], cloud computing
scaling. has evolved to a new generation referred to as server-
less computing. However, serverless was not a brand-new
Decrease latency paradigm; it emerged after the advancements in adopt-
Serverless applications are not hosted on a specific server; ing virtual machines and container technologies [120]. By
the code can run from any server in any location. There- 2016, other competitors, namely Google, Microsoft, and
fore, cloud vendors can run the application on servers near IBM followed the trend. Several commercial and open-
the end user’ location. This reduces latency, because end source platforms offer serverless computing. The well-

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 11 of 29

Table 4 Organizations, and researchers active in researching serverless computing

Organization Author(s) Published Paper Total
IBM Research, United States Alaa Youssef, Aleksander Slominski, Ali Anwar, Ali Kanso, Animesh [17, 18, 21, 35–54] 23
Trivedi, Bruce Slawson, Canturk Isci, Dan Williams, Dimitrios Skourtis,
Eric Rozner, Erik Wittern, Ioana Baldini, Jonas Pfefferle, Kerry Shih-Ping
Chang, Lionel Villard, Lukas Rupprecht, Michael Maximilien, Nick
Mitchell, Nilton Bila, Nima Kaviani, Nirmal K Mukhi, Olivier Tardieu, Paolo
Dettori, Patrick Stuedi, Paul Castro, Pedro Garcáa López, Perry Cheng,
Philippe Suter, Prabhakar Kudva, Ricardo Koller, Rodric Rabbah, Shripad
Nadgowda, Srijith Prabhu, Stephen Fink, Vasily Tarasov, Vatche Ishakian,
Vinod Muthusamy, Waldemar Hummer
Technische Universität Berlin, Jorn Kuhlenkamp, Sebastian Werner, Maria C. Borges, Stefan Tai, [13, 20, 55–62] 10
Germany Johannes Müller, Richard Girke, Jörn Kuhlenkamp, Dominik Ernst, Frank
Pallas, Niklas Anders, Nebi Mucaj, Olesia Tsaplina, Christian Schmidt,
Kann Yildirim, Tobias Pfandzelter, David Bermbach, Ahmet-Serdar
Karakaya, Simon Buchholz, Setareh Maghsudi, Jonathan Hasenburg
Universitat Rovira i Virgili,Spain Pedro Garcáa López, Marc Sánchez-Artigas, Gerard Pará, Daniel [45, 46, 49–51, 63–66] 9
Barcelona Pons, Álvaro Ruiz, David Arroyo Pinto, Josep Sampé, Amanda
Gómez-Gómez, Mariano Ezequiel Mirabelli, Germán T. Eizaguirre, Aitor
Arjona
University of California, Berkeley, Andrew Zhang, Benjamin Recht, Chenggang Wu, Dominic Carrano, Ion [67–74] 8
United States Stoica, Jeffrey Ichnowski, Joao Carreira, Joseph E. Gonzalez, Joseph M.
Hellerstein, Kannan Ramchandran, Karl Krauth, Ken Goldberg, Qifan Pu,
Randy Katz, Samuel Paradis, Saurav Chhatrapati, Scott Shenker, Thomas
Courtade, Vaishaal Shankar, Vikram Sreekanti, Vipul Gupta, Vivian Fang,
Wen Zhang, Yaoqing Yang
University of Washington, Tacoma, Baojia Zhang, David Foster, David Perez, Derek Chen, Dimitar Kumanov, [75–82] 8
United States Eyhab Al-Masri, Hanfei Yu, Ibrahim Diabate, Ka Yee Yeung, Ling-Hong
Hung, Ming Hoi Lam, Minh Vu, Mohammadbagher Fotouhi, Rashad
Hatchett, Richa Jain, Robert Cordingly, Shruti Ramesh, Swetha
Chinthalapati, Swetha Reddy Nathala, Varik Hoang, Wen Shu, Wes
Lloyd, Xingzhi Niu, Zohreh Sadeghi
Vrije Universiteit Amsterdam, Lucian Toader, Alexandru Uta, Ahmed Musaafir, Alexandru Iosup, [16, 83–89] 8
Netherlands Laurens Versluis, Erwin Van Eyk
University of Wisconsin-Madison, Aditya Akella, Andrea C. Arpaci-Dusseau, Arjun Singhvi, Dennis Zhou, [71, 74, 90–94] 7
United States Edward Oakes, Kevin Houck, Leon Yang, Liang Wang, Michael Swift,
Remzi H. Arpaci-Dusseau, Scott Hendrickson, Shivaram Venkataraman,
Stephen Sturdevant, Tyler Harter, Venkateshwaran Venkataramani
Delft University of Technology, Erwin Van Eyk, Alexandru Iosup, Sacheendra Talluri [16, 84–86, 89, 95] 6
Netherlands
Escuela Superior Politecnica del Gabriel Aumala, Edwin F. Boza, Luis Ortiz-Avilés, Gustavo Totoy, Cristina [2, 85–87, 95, 96] 6
Litoral, Ecuador L. Abad, Mónica Villavicencio, Stephany Quimba
Stanford University, United States Ana Klimovic, Yawen Wang, Christos Kozyrakis, Sadjad Fouladi, Keith [43, 53, 71, 97–99] 6
Winstein, Kostis Kaffes, Neeraja J. Yadwadkar, Sean Choi, Muhammad
Shahbaz, Balaji Prabhakar, Mendel Rosenblum, Kailas Vodrahalli
University of Bamberg, Germany Johannes Manner, Martin Endreß, Tobias Heckel, Guido Wirtz, Stefan [1, 100–104] 6
Kolb, Stefan Winzinger, Marcel Großmann, Christos Ioannidis, Duy
Thanh Le, Mike Prechtl, Robin Lichtenthäler
Imperial College London, United Robert Chatley, Alim Ul Gias, Giuliano Casale, Alessandra Russo, Lulai [8, 105–108] 5
Kingdom Zhu, Jessica Vandebon, José G. F. Coutinho, Wayne Luk
Kookmin University, South Korea Jeongchul Kim, Kyungyong Lee, Jungae Park, Hyunjune Kim, Jaeghang [109–113] 5
Choi
Shanghai Jiao Tong University, Mingyu Wu, Zeyu Mi, Yubin Xia, Zijun Li, Quan Chen, Shuai Xue, Minyi [114–118] 5
China Guo, Dong Du, Tianyi Yu, Binyu Zang, Haibo Chen, Qingyuan Liu, Ziqian
Lu, Jinfeng Liu, Zheng Huang, Zhichao Hua
Universitat Politècnica de València, Alfonso Pérez, Germán Moltó, Miguel Caballer, Amanda Calatrava, [22, 119–122] 5
Spain Sebastián Risco, Diana M. Naranjo, V. Giménez-Alventosa, Carlos De
Alfonso, Ignacio Blanquer

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 12 of 29

Table 4 Organizations, and researchers active in researching serverless computing (Continued)

Organization Author(s) Published Paper Total
University of Alberta, Canada Di Niu, Sara Ghaemi, Alireza Goli, Omid Hajihassani, Omid Ardakanian, [123–127] 5
Moe Rashidi, Tyler Dauphinee, Nima Mahmoudi
University of Stuttgart, Germany Michael Wurster, Uwe Breitenbücher, Kálmán Képes, Frank Leymann, [11, 107, 128–130] 5
Vladimir Yussupov, Michael Hahn, Christian Müller, André van Hoorn
Zurich University of Applied Josef Spillner, Seán Murphy, William Martini, Bill Bosshard [17, 131–134] 5
Sciences, Switzerland
IBM Research, Israel Alex Glikson, Gil Vernik [23, 49, 65, 66] 4
Intel Corporation, United States Michael Steiner , Ramesh G. Illikkal , Eriko Nurvitadhi , Mishali Naik , [108, 135–137] 4
Anup Mohan , Harshad Sane , Kshitij Doshi , Saikrishna Edupuganti ,
Naren Nayak
Microsoft Research, United States Lan Ly, Mohammad Shahrad, Rodrigo Fonseca,Íñigo Goiri, Gohar [68, 75, 91, 138] 4
Chaudhry, Paul Batum, Jason Cooke, Eduardo Laureano, Colby Tresness,
Mark Russinovich, Ricardo Bianchini, Tyler Harter, Jose M. Faleiro
Platform9 Inc., United States Erwin Van Eyk [84, 85, 89, 95] 4
Rensselaer Polytechnic Institute, Matthew Obetz, Anirban Das, Timothy Castiglia, Stacy Patterson, Ana [139–142] 4
United States Milanova, Shigeru Imai, Andrew Leaf, Carlos A. Varela
University of California, Santa Wei-Tsung Lin, Chandra Krintz, Rich Wolski, Michael Zhang, Fatih Bakir, [10, 143–145] 4
Barbara, United States Gareth George
University of Illinois, Mengting Yan, Pubali Datta, Adam Bates, Arnav Sankaran, Tarek Elgamal [44, 146–148] 4
Urbana-Champaign, United States
University of Tartu, Estonia Chinmaya Kumar Dehury, Satish Narayana Srirama, Tek Raj Chhetri, [107, 149–151] 4
Pelle Jakovits
University of Würzburg, Germany Johannes Grohmann, Simon Eismann, André Bauer, Norbert Schmitt, [85–88] 4
Nikolas Herbst, Samuel Kounev
York University, Canada Hamzeh Khazaei [124–127] 4
AGH University of Science and Maciej Malawski, Adam Gajek, Adam Zima, Bartosz Balis, Kamil Figiela, [152–154] 3
Technology, Poland Piotr Moczurad, Maciej Pawlik, Pawel Banach
Boston University, United States Anthony Byrne, Ayse K. Coskun, Nabeel Akhtar, Ali Raza, Ibrahim Matta, [48, 155, 156] 3
James Cadden, Thomas Unger, Yara Awad, Han Dong, Orran Krieger,
Jonathan Appavoo
Budapest University of Technology David Balla, Markosz Maliosz, Csaba Simon, János Czentye, Balázs [157–159] 3
and Economics, Hungary Sonkoly
Einstein Center Digital Future Tobias Pfandzelter, David Bermbach, Ahmet-Serdar Karakaya, Simon [60–62] 3
Mobile Cloud Computing Research Buchholz, Setareh Maghsudi, Jonathan Hasenburg
Group, Germany
George Mason University, United Benjamin Carver, Jingyuan Zhang, Ao Wang, Yue Cheng [52, 54, 160] 3
States
George Washington University, Guyue Liu, Timothy Wood, Phani Kishore Gadepalli, Sean Mcbride, [6, 161, 162] 3
United States Gregor Peach, Gabriel Parmer
Google Inc., United States Qifan Pu, Junaid Khalid, Vadim Sukhomlinov [71, 93, 137] 3
Indian Institute of Technology Nikhila Somu, Nilanjan Daw, Umesh Bellur, Purushottam Kulkarni, S. C. [163–165] 3
Bombay, India Gupta
Macquarie University, Australia Young Choon Lee [166–168] 3
Stony Brook University, United Amoghavarsha Suresh, Gagan Somashekar, Anandh Varadarajan, [146, 169, 170] 3
States Veerendra Ramesh Kakarla, Hima Upadhyay, Anshul Gandhi, Prabuddha
Kumar, Amir Rahmati
University of Colorado Boulder, Zaid Al-Ali, Sepideh Goodarzy, Ethan Hunter, Sangtae Ha, Richard Han, [21, 171, 172] 3
United States Eric Keller, Erika Hunhoff, Shazal Irshad, Ali Tariq, Eric Rozner, Austin
Pahl, Sharat Nimmagadda, Siddharth Lanka
University of Gothenburg, Sweden Philipp Leitner, Joel Scheuner [17, 30, 87] 3
University of Sydney, Australia Mohammadreza Hoseinyfarahabady, Albert Y. Zomaya, Young Ki Kim [166–168] 3
University of Waterloo, Canada Matt Crane, Jimmy Lin, Youngbin Kim, N. Asokan [135, 173, 174] 3

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 13 of 29

Fig. 8 Active countries

known commercial systems are AWS Lambda, Google programming languages and model, deployment easiness,
Cloud Functions, and Azure functions. Alternately, there easiness in composing functions from different providers,
are several open source platforms available including IBM security, and monitoring and debugging tools [184].
Cloud Functions, and Apache OpenWhisk. Table 7 presents the serverless platforms used in the
There are several criteria to help developers in select- considered papers of this study. It can be noted that “AWS
ing a serverless platform: cost, performance, supported Lambda”, “Apache OpenWhisk”, and “Azure Functions” are
the most used platforms with 78, 23, and 11 published
papers, respectively. However, it is worth mentioning that
Table 5 Comparison between serverless and traditional each platform has its own set of features and differs from
computing others.
Factors Traditional cloud Serverless
computing computing The application areas of serverless computing in the
Development phase Difficult Easy literature (RQ6)
Automatic scalability Unavailable Available Serverless computing can be utilized in a number of
Stateful applications Easy Difficult application areas as follows:
Security Complex and less Easy and more secure Chatbot
secure
A chatbot application is developed using serverless com-
Functions’ life cycle Long Short
puting, which enables interaction with users via voice
Troubleshooting and Easy Difficult
debugging
Server and hardware Required Unavailable
configuration and Table 6 Benefits of serverless computing
maintenance Benefits Papers Total
Failure tolerance Less reliable More reliable
Cost effective [16, 94, 134, 180–190] 14
Cost (Variable Expensive Affordable
Scalability [15, 16, 74, 75, 90, 94, 153, 188, 12
workload)
189, 191–193]
Cost (Stable Affordable Expensive
Server-side management [15, 16, 121, 181, 188, 194–197] 9
workload)
Easy to deploy [6, 15, 46, 95, 97, 131, 189] 7
Applicable user Administrator and Developer
developer Decrease latency [9, 95, 179, 198] 4

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 14 of 29

Table 7 Serverless platforms usage are implemented as Amazon lambda functions. The
Platforms Papers Total search engine executes all the details of retrieval pro-
AWS Lambda [3, 5, 18–20, 30, 37, 46, 55, 68, 78 cessing after receiving the user query (e.g., tokenization,
69, 75, 76, 78–80, 83, 91, 100, stop-word removal, term weighting, similarity calculation,
101, 106, 109, 110, 119–121,
123, 129, 130, 140, 143, 150,
and ranking). Then, it sends back the results to the user
152, 153, 158, 160, 166, 167, as documents stored in the DynamoDB database to be
173, 174, 176, 177, 179, 181, accessed using the web application interface [173].
183, 188, 189, 200–230]
File processing
Apache OpenWhisk [35, 36, 38, 41, 44, 91, 135, 138, 23
183, 188, 189, 206, 208, Serverless computing can been utilized in file process-
231–240] ing applications [119, 249]. For instance, in [119] a model
Azure Functions [55, 75, 77, 100, 109, 130, 183, 11 for highly parallel file processing applications based on
204, 206, 208, 230] serverless architecture is proposed. This model provides
Google Cloud Functions [79, 109, 152, 183, 206, 208, 7 users with different ways to process their files.
241] The first method is by using the API gateway. In
OpenFaaS [22, 122, 141, 235, 242–244] 7 this method, users submit files using the HTTP request
OpenLambda [90–92, 96, 245, 246] 6 employing the API gateway to a lambda function to pro-
IBM Cloud Functions [65, 130, 183, 188] 4 cess the file (e.g., medical images and video files).
Knative [235, 243] 2
The second method is by uploading/reading files to the
Amazon simple storage service (Amazon S3) bucket. This
FunctionStage [247] 1
method provides the user with three different ways to exe-
Nuclio [243] 1 cute a lambda function using S3 buckets: (a) by uploading
a file to S3 buckets. When the file is uploaded, S3 creates
an event to invoke a lambda function; (b) by copying a file
or text commands. Within this application, six function- from another bucket to the bucket linked with the lambda
alities have been implemented, namely the Date, News, function. This will cause the trigger of an event from S3
Jokes, Weather, Music Tutor, and Alarm Service. For to invoke a lambda function as in the previous manner; (c)
example, a user can ask for the current date using a voice by specifying a bucket where the files to be processed are
or text command. The request is routed to a required stored. Then, for each file found, the lambda function is
serverless action on OpenWhisk for further processing. invoked in parallel using an S3 bucket.
The Date action is activated via the issued command and The third method is by specifying the output file. By
retrieves the current date to the user in the form of text or this method, the user can set a chain of lambda functions
voice [44]. to be invoked by S3 buckets. In this case, the user defines
Another example is the ticketing chatbot service devel- the input/output buckets for each of the lambda func-
oped using serverless computing and natural language tions. Thus, the output bucket can be used as an input to
processing (NLP). The architecture of the system con- another lambda function [119]. Here, serverless functions
sists of three parts: (1) the node.js Webhook, which works can handle different types of data (stored in files) such
based on hypertext transfer protocol (HTTP) POST or as sensory, textual, and biological data [200]. Also, many
GET requests (2) Wit.AI, which is a NLP service (3) preprocessing operations using NLP may be applied to
Ticket.com, which is a ticketing order API. For example, data files before processing, such as stemming and noise
when a user books a flight ticket; a specific function on removal [78].
Webhook will be activated, which routes the user query
to the Wit.AI service. Wit.AI will process the query and Smart grid
extract useful parameters from the request such as des- A MATLAB simulation scenario is created to illustrate
tination, date, and time, then send it back to Webhook. the use of the smart grid with serverless cloud comput-
After receiving the processed query from Wit.Ai; another ing to control and manage electrical loads (devices). In
action will be triggered and pass the processed query to this scenario, the Simulink tool is employed for simula-
Tickt.com API to retrieve flight information such as the tion. A MATLAB program is developed to indicate the
flight number, airline name, departure time, and ticket start and end of the simulated grid model via a batch file.
price from several airline companies. Finally, Webhook The batch file is used to upload grid model data gener-
will provide flight information to the user [44, 179, 248]. ated by the program to Amazon S3. Afterwards, a lambda
function in the serverless side will be activated to process
Information retrieval the uploaded data, and subsequently the result will be sent
A search engine web-based application is developed based back to the batch file as a response. In return, the program
on serverless architecture. Search engine functionalities will read continuously the response from the batch file

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 15 of 29

and interpret its content to manage the electrical switch SDN is a network architecture approach that enables the
(loads) [201]. network to be manageable and adaptive. This architecture
Also, An electrical overload warning system is imple- separates the network control plane from the forwarding
mented in the smart grid, based on serverless architecture. functions (the data plane). This decoupling enables net-
The system uses some Amazon web services, including work switches to become a simple forwarding device, and
S3, lambda functions, simple notification service (SNS), the network control is implemented as a network appli-
and CloudWatch. S3 is used as a storage service in the cation that executed on a logically centralized controller.
system. Lambda functions constitute a computing service Serverless computing can be used in the SDN controllers.
that executes the code of the application. CloudWatch is a These controllers can be implemented as independent
monitoring tool that monitors AWS resources and appli- functions deployed on serverless platforms. For example,
cations. The SNS is a notification service that sends and when a packet arrives to the SDN forwarding device, the
receives notifications. device will parse the packet header and forward it to the
The main sections of this warning system consist of SDN controller. The functions within the SDN controller
data collection, data acquisition, data analysis, data min- will be activated then it will determine what action to be
ing, conclusion verification, and conclusion publishing. In taken with the packet. After that, it will send the infor-
this architecture, the AWS Lambda is used in data analysis mation to the forwarding device. The action might be
and data mining. AWS CloudWatch is used for data con- modifying the header, dropping the packet, etc.
clusion verification. The SNS is used to generate alarms.
For instance, the data is uploaded to S3, and subsequently, IoT
a lambda function is activated for data analysis and data Serverless computing has been utilized in many IoT appli-
mining. After the lambda function execution, its log data cations, as shown in Table 8. For example, a camera can
is stored in CloudWatch logs. CloudWatch is used for con- be installed to monitor a house, after which processing
clusion verification. CloudWatch defines an alarm size to images captured by the camera can be performed by some
a specific value, upon which it compares the value of log serverless functions provided by the OpenWhisk plat-
data with a predefined alarm size to check the current form. When a camera detects an interesting object such
state. Then, the CloudWatch uses SNS for publishing con- as a car or a human, the camera sends its pictures to the
clusions. If the receiving data is greater than the alarm serverless platform for further processing. To extract fea-
size, an alarm signal will be triggered and send an email tures, a serverless function is called to perform feature
via SNS [5]. extraction and then reports its status to the users [232].

Security Edge computing

An automated threat detection system is introduced using Serverless cloud computing and edge computing have
serverless cloud computing and Kubernetes. Kubernetes been used to build different kinds of applications, as
is an open source system to automatically deploy and presented in Table 8. For instance, the authors of [217]
manage application containers [243, 250]. The system have implemented an autonomous mobile robot (AMR)
deals with threats (e.g., software vulnerabilities and inse- system based on serverless computing and edge comput-
cure configurations) automatically based on user-defined ing. The system consists of three main components: an
policies. The system includes a vulnerability scanner (VS), AMR with NVIDIA Jetson TX2 module for edge com-
which is a thread detection component. Whenever users puting, a serverless architecture based on AWS, and a
deploy new application containers, the containers are reg- cross-platform mobile application developed using React
istered with the VS, and a scanner agent is installed. Native. The main idea of the system is to deliver a pack-
When a thread is detected by the scanner, a notification age to a user. For example, the user will interact with the
is sent to the OpenWhisk component, which activates a
serverless function that takes actions to reduce the threat.
OpenWhisk will invoke a Kubernetes API extension and Table 8 Top three serverless application areas
let the security enforcement operator (SEO) handle the Application Area Papers Total
operation [35]. Edge Computing [6, 23, 60, 62, 140, 144, 145, 148, 25
151, 158, 161, 162, 189, 193, 217,
Networks 222, 231, 232, 235, 237, 253–257]
Serverless cloud computing has been employed in dif- Security [12, 35, 94, 97, 104, 114, 124, 135, 20
ferent networking domains[175, 188, 251, 252]. In [188], 146, 147, 176, 178, 187, 221, 256,
258–262]
a variety of networking fields including software-defined
networking (SDN) which can utilize advantages of server- IoT [1, 10, 60, 77, 144, 151, 158, 162, 19
222, 232, 234, 237, 241, 255–257,
less computing architecture have been discussed. The 260, 263, 264]

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 16 of 29

mobile application to send a package. Once the delivery includes CPU usage, memory, execution time, and band-
request has been received from the user, the AWS IoT can width [94, 202, 210, 235, 280].
activate related lambda functions, such as position coor-
dinate. Then, the AMR would start its mission, sending Security
the package to the receiver’s location. Also, facial images Security is the most challenging issue in serverless cloud
were regularly retrieved by AWS lambda to identify the computing. One of the security issues is isolation, because
receiver’s face. Finally, the task is completed when the functions are running on a shared platform by many users.
receiver’s identification is confirmed [217]. Therefore, strong isolation is required. Another security
issue is trust when it comes to process-sensitive data. The
Serverless computing challenges and issues (RQ7) serverless applications work with many system compo-
Studying the literature reveals a number of challenges nents, which must function correctly to maintain security
and issues posed by employing serverless computing. properties. Table 9 presents several papers associated with
These challenges cover the functional and non-functional serverless security.
aspects of serverless computing as follows:
Scalability
Cost and pricing model Serverless computing must ensure function scalability
Cost is a fundamental challenge; therefore, serverless and elasticity. For example, when many requests are
computing providers should reduce the usage of resources issued to a serverless application, these requests should
to the minimum, while functioning in both execution and all be served and the used serverless cloud provider
idle states. Further, the pricing model is another challenge should provide the required resources to process all these
in serverless computing compared to other cloud com- requests and should scale up with the number of requests
puting approaches. For example, the CPU bound is cheap, [210, 280, 281].
whereas the input/output (I/O) bound functions may be
Long-running
more expensive from dedicated servers. Table 9 presents
Serverless computing runs function in a limited and short
papers that investigate issues on cost and pricing models
execution time, while there are some tasks might require
in serverless cloud computing.
long execution time. This does not support long execution
Cold start running, since these functions are stateless, which means
Serverless computing can scale to zero while there is no that if the function is paused it cannot be resumed again
request for functions and services. Scaling to zero leads [11, 202, 234, 280].
to a problem called cold start. A cold start occurs when
Programming & debugging
serverless functions remain idle for some time, and the
There is currently a lack of debugging tools. Further,
next time these functions are invoked, a longer start time
monitoring tools are required, since developers need to
is required. Methods and techniques to reduce the cold
monitor the application and observe how functions are
start problem are crucial as a result, many papers have
working. More advanced integrated development envi-
been studied this problem, as shown in Table 9.
ronments (IDEs) are needed, so developers can perform
Resource limits refactoring functions, such as merging or splitting func-
In serverless computing, resources are required to ensure tions, and reverting functions to the previous version.
that the platform can deal with load increasing. This Moreover, logs from serverless function invocations need
to be sent to the developer and provide detailed stack
Table 9 Top challenges and issues traces. When an error occurs, a good method is required
Challenges and Issues Paper Total to report details on the occurrence to the developer.
Performance [11, 16–18, 44, 81, 84, 85, 33 The equivalent of a stack trace for serverless comput-
89–91, 95, 96, 114, 126, 157, ing is currently not available. Table 9 shows many papers
161, 162, 203, 204, 210, 219,
226, 227, 232, 262, 265–271] that consider programming and debugging challenges and
issues.
Cold start [91, 98, 100, 105, 137, 149, 172, 17
202–204, 210, 229, 232,
272–274, 280] Vendor lock-in
The FaaS paradigm separates the code from the data,
Security [11, 17, 44, 84, 97, 98, 114, 187, 13
210, 256, 258, 275, 280] which leads the functions to depend strongly on the could
Programming and [9, 11, 15, 16, 44, 101, 119, 120, 13
provider’s ecosystem for storing, obtaining, and transfer-
debugging 210, 235, 276, 277, 280] ring data [282]. This issue makes the customers dependent
Cost and pricing [11, 16, 81, 88, 202, 204, 210, 11 on the serverless provider for products and services, and
267, 278–280] the customers cannot easily use different vendors in the

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 17 of 29

future without substantial cost. Thus, customers have to Legacy systems

wait on the serverless provider for additional services Legacy systems refer to old technologies, techniques,
[9, 130, 202]. hardware, and software systems that are still in use. It
should be possible to reach these systems from server-
Performance less cloud computing. Also, these systems might be
Serverless computing has many performance challenges required to be transferred to cloud computing. There-
and issues such as scheduling and service calling over- fore, more work needs to be done on the migration pro-
head. For instance, scheduling means when a serverless cess and how the functions can be extracted from legacy
function is activated in response to an event this function systems to be deployed as serverless cloud functions
should be mapped to a specific resource (e.g., container or [84, 119, 120, 210, 280].
VM) to be run. The resource can have a significant effect
on performance based on available resources, location of Managing hybrid cloud
input data and code, load balancing, etc. Table 9 shows In a hybrid cloud, a developer may deploy an applica-
papers related to serverless performance. tion to different clouds (hybrid cloud). For example, if
some functions of an application are on a specific server-
Fault tolerance less cloud vendor and others are hosted on other public
It refers to a system that continues working and pro- clouds; then, managing these functions and their interac-
vides its services despite the failure in some components. tions is a challenging issue [84, 210, 280].
It mostly occurs when some containers fail. To over-
come this challenge, a basic retry mechanism is used Lack of quality of service (QoS) support
[11, 210, 235]. Existing serverless platforms and frameworks do not pro-
vide users the control over the QoS of serverless functions
Function composition [235]. Cold starts, queuing, and orchestration are the main
Serverless cloud vendors provide users the ability to reasons affecting the QoS in serverless computing [8].
deploy small stateless functions to the cloud to handle a
specific task. However, some complex tasks require mul- Architecture complexity
tiple functions to work with each other collaboratively to A serverless application may consist of several func-
be performed. Therefore, more research needs to be done tions; the number of functions increases the complexity
on how function composition can be used effectively and of the architecture. Managing these functions and ser-
efficiently in serverless cloud computing [11, 38, 235]. vices related to the application also leads to a complex
architecture [9].
Resource sharing
Functions in serverless cloud computing share resources Interactions tracking
to achieve inexpensive cloud computing. Sharing Stateful requests are usually used by real-life applications.
resources among functions and other serverless compo- It means deployed systems keep track of the state of users’
nents is a challenging task. Therefore, good techniques interactions and store them on the server-side for further
are required to be investigated to achieve this goal uses. However, in stateless serverless functions, it is not
[98, 210, 283]. obvious how these functions will handle and manage the
states of each user [210, 280].
Testing
A serverless application consists of many small functions. Concurrency management
These functions work together to accomplish the appli- Concurrency means a function can handle any number
cation’s functionality. Therefore, integration testing for of requests whenever a function is invoked. For exam-
these functions is a crucial issue to make sure that the ple, if a request has been made to a serverless function,
application works properly [9, 84, 284]. the function will process that request. However, if another
request has been made to that function and the function
Naming and addressing system is still processing the previous request, then the serverless
Users deploy functions to serverless cloud computing to should provide another instance of that function to serve
solve problems. These functions cannot listen to network the new request [210, 280].
communications. The existing serverless cloud comput-
ing frameworks do not support this feature. Instead, they Support for heterogeneous hardware
use third party services such as Amazon S3 to communi- Existing serverless platforms may not support some spe-
cate with other functions. Therefore, finding the internet cialized hardware such as graphics processing unit (GPU)
protocol (IP) address of a function by other functions and field programmable gate arrays (FPGAs). This is a
and services is a challenging issue in serverless cloud challenging issue for vendors to provide support for het-
computing [98]. erogeneous hardware [210, 280].

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 18 of 29

Tools available for serverless computing (RQ8) integrating the latest technological development into
Nowadays, various providers strive to facilitate the monolithic systems is a tough and expensive process. Gen-
adjustable use and allocation of machine resources on the erally, monolithic applications are designed to be tightly
cloud [9]. Likewise, plenty of supportive tools and ser- coupled – the entire application will be unable to run
vices are aiding developers to more efficiently manage and or compile if one component is missing or fails [128].
deploy applications using serverless computing. Server- It is also difficult to scale the application when multiple
less computing is auto-scalable, reliable, and easily acces- components have limited resources.
sible [203]; for these reasons, big cloud providers such Another drawback is that updating any component will
as Amazon, Microsoft, Google, IBM have realized the require redeployment of the entire project. The migration
importance of offering frameworks, IDEs, software devel- process to serverless computing involves transferring the
opment kits (SDKs), function development kits (FDK), legacy application code to serverless functions. This pro-
migrating mechanisms, logs, and monitoring tools to cess could be more efficient and functional in applications
enhance and simplify the development, testing, deploy- with less size [76].
ment, and monitoring of serverless applications [17]. For The key challenging aspect of migration is about
instance, Amazon offers Cloud9 IDE for local deploying extracting the serverless computing from the monolithic
and testing [205]. systems. There are several approaches to accomplish this
Apart from the cloud providers’ specific tools, plenty task, one of them is Lift and shift [205]. This technique
of third-party tools exist for the developers. With the transfers the whole infrastructure to the cloud, however,
concept of these tools, developers can build and deploy this method also brings the already existing problems
applications on multi-cloud providers. Developers are within the source to the destination. In [205] the authors
also able to control platforms and resources by program- proposed toLambda to automatically refactor, test, and
ming. The advantages of this are linking the applications deploy the monolithic applications (Java) into microser-
with auto-scaling controllers and including advanced self- vices (AWS Lambda Node.js). While rebuilding the legacy
mechanisms into the code to automatically configure, application from scratch is recommended for applications
secure, optimize, and recover the cloud applications. The that no longer depend on the existing cloud services [130].
core advantage of this feature is the acceleration in apply- However, not all applications are suitable for migra-
ing changes to the application environment [272]. tion to serverless computing [76, 128]; therefore, the first
There are several tools available to model serverless important aspect to be considered before rebuilding the
applications, which are based on deployment models as applications is whether it would save money [188]. For
either imperative or declarative. The imperative model such cases, newly desired features could be implemented
defines the execution steps to obtain a specific deploy- and added via serverless computing as an extension to the
ment task. While the declarative model describes the current systems [128].
structure of a desired application deployment. However, The other approach is to refactor the entire legacy code
to fully benefit from employing a serverless architecture, into FaaS services. During the migration phase, it is cru-
cloud providers should address issues that have arisen cial to address the coupling of the systems not only in the
with the use of a serverless paradigm. For instance, debug- application logic but also in the databases, as more func-
ging tools are unable to track and identify the exact reason tions will call the same database. However, migrating the
behind errors [44], as most of them are limited to what server-side while keeping the user interface could lead to
cloud providers offer [179]. Although many powerful tools problems. Moreover, the client cannot obtain integrated
have been mentioned in this study and can be used in data by a single request. As the functions are decoupled
serverless computing in real scenarios, there is still a great into smaller entities, the server is unable to aggregate data
opportunity to develop further tools and services. from different entities. Thus, it is the client’s responsibility
to call the necessary entities to achieve this task [76].
Migration of monolithic applications to serverless
computing (RQ9) Future directions of research (RQ10)
The nature of most existing applications is monolithic. As the evolution of serverless computing is relatively new,
Monolithic applications have several drawbacks; they are there are several research paths available to be focused on
characterized by continuous growth in complexity and as follows:
size over time.
The bigger size of the monolithic applications leads Function startup
to slower startup time. Moreover, novice developers One of the major research opportunities is overcom-
face difficulties in digesting the traditional program- ing the cold start problem without affecting the pri-
ming paradigm. Economically, monolithic systems take mary feature of serverless which is scaling to zero
more effort to be developed and debugged. Furthermore, [160, 188]. The first call of functions needs initializing

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 19 of 29

the required libraries, which will cause a cold start. To Pricing

bypass this, the computing resources will be warm for a Pricing is crucial for both customers and cloud providers.
certain time. Hence, upcoming requests will be handled However, there is a shortage in pricing models, as there
faster. This could be performed via enhancing scheduling is an imbalance in needs between serverless providers,
policies and developing more accurate function perfor- developers, and service end-users [236]. The pricing
mance measurements [86]. Serverless providers follow scheme for most cloud providers is based on the num-
their approaches to keep the functions in the warm pool. ber of functions’ requests and execution time-the quantity
However, most of them are based on the number of of consumed resources [123, 200]. Currently, FaaS is less
requests for a certain time. Thus, if a function is not called expensive when functions are bound to I/O than CPU.
frequently, it will suffer again from the cold start. Moreover, services that dynamically adjust resource con-
Very few studies such as [272] suggested a periodic sumption are unable to predict the optical computing
event scheduler for Kotless (a serverless framework for technology. It is crucial to implement solutions that offer
Kotlin) which will trigger a list of warm functions every cost-effective computing resources. FaStest reduced the
few minutes. The authors of the study claimed that this cost by 50% via learning the behavioral pattern of func-
will reduce the cold start without bringing extra costs. tions using machine learning [202]. Price estimation has
While in [233] argues that pre-warming methods are a great impact on selecting the most optimal provider.
unnecessarily using resources with idle containers. The Therefore there should be more researches on developing
researchers are still working to avoid cold start by reduc- tools to predict the pricing in advance.
ing high delayed function startups via optimizing compute
resources [11]. Legacy systems
Recycling and rebalancing minutes and hours of idle Since the serverless emergence, researchers are work-
runtime is an expensive process for cloud providers. ing on the open question of how to decompose legacy
Therefore, reducing the cold start penalties will help systems into FaaS without degrading performance [208].
cloud providers in the first place and hence customers. Several works have been done on migrating to FaaS
The authors in [202] proposed FaaStest an autonomous [76, 130, 286]. The currently available automated tools
approach based on machine learning to capture the func- for migrating legacy code into FaaS are not fully practical
tion call behavior and then dynamically select the optimal due to the remaining manual work that needs to be done
ones. This technique could reduce the cold start by 90%. [17]. Therefore, finding optimal automatic migration solu-
They proposed a strategy to predict functions invoking tions for existing legacy systems is an interesting research
time and warming the function using fine-grained regres- direction [130]. Moreover, research on tools for checking
sion method [285]. However, overcoming the issue of whether a legacy system will fit the serverless paradigm is
function startups is still considered as a research direction a crucial line. Also, developing and enhancing automatic
to be more investigated. and semi-automatic analysis strategies based on artificial
intelligence could be another future research field.
QoS
Keeping a guaranteed QoS level in the software level Debugging, testing, and benchmarking
agreement (SLA) that describes the lower service level The available tools for testing, debugging, and deployment
offered by the service providers [166] is a major obstacle are immature, this prevents some developers from enter-
for cloud providers to offer optimal performance met- ing the serverless environment. The shortage of tools in
rics [167, 207]. However, serverless frameworks should FaaS is a core problem, particularly the testing tools [17].
consider the objectives of both providers and users Moreover, most FaaS environments lack powerful local
[242]; customers and developers have none or little emulation platforms for testing. Therefore, developers are
QoS support over the functions [236]. In addition, the mostly depending on the server-side, which is expensive.
auto-scaling feature lacks QoS guarantees. This lack of Developers need to be ensured about the adequate testing
QoS affects the performance of serverless applications. tools before diving into the serverless world. A challeng-
Increasing response time leads to decreasing the QoS ing aspect in benchmarking is the lack of information due
level [207]. It also raises the cost of the service [236]. to the heterogeneity of the cloud provider data center:
Therefore, achieving an ideal resource allocation man- hardware, software, and configurations [287]. Addition-
agement is a complicated and challenging task as sev- ally, benchmarking FaaS platforms should take advantage
eral objectives should be fulfilled together [209]. Hence, of analyzing the cloud services, which lacks limited acces-
providing more efficient QoS management of functions sible measurements and hidden modification of services
by the auto-scaling is essential to be considered with- over time [55]. Thus, it is essential to have transparent,
out degrading the fault-tolerance features and increasing fair, and standardized benchmarking tools available for
the cost. developers.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 20 of 29

Threats to validity in further investigating serverless computing. Finally, it is

Several threats might impact the validity of the litera- worth mentioning that the interest that both commercial
ture mapping studies. In this paper, popular instructions and academic efforts fueled into studying, developing, and
and guidelines were taken into account to avoid threats to implementing serverless tools in forthcoming years could
validity as follows: help maximize the potential that serverless computing
could bring to the IT community.
• Coverage of research questions: All up-to-date
Acknowledgements
research aspects of serverless computing might not
Not applicable.
be included in this study. To overcome this threat,
the brainstorming was conducted by all the authors Authors’ contributions
Conceptualization: HBH, SAB, and QIS; methodology: HBH, SAB, and QIS;
in determining the most current research questions
validation: HBH, SAB, and QIS; formal analysis: HBH, SAB, and QIS; investigation:
in the area. HBH, SAB, and QIS; resources: HBH; data curation, HBH and SAB;
• Coverage of related papers: The process of obtaining writing—original draft preparation: HBH, SAB, and QIS; writing—review and
all the related studies in serverless computing cannot editing: HBH, SAB, and QIS; visualization: SAB; supervision: QIS; It is noted that
all authors cooperated with each other to achieve suitable information flow
be secured. In this study, various literature databases across the entire paper. The authors read and approved the final manuscript.
were employed; moreover, the method based on
different terms and synonymous is followed by all the Authors’ information
Hassan B. Hassan received the B.Sc. degree in Computer Science from
authors in determining the related questions. University of Duhok, Iraq, in 2010. He completed the M.Sc. degree in Web
• Paper inclusion/exclusion criteria: The individual Applications and Services, from Leicester University, UK, in 2015. He is currently
bias and interpretation could affect the working as an assistant lecturer at the college of medicine, University of
Duhok, Iraq. His main areas of research interest are cloud computing, web
implementation of the criteria. Therefore, to solve programming, big data, and human computer interaction.
this problem, the agreements of all authors were Saman A. Barakat received the B.Sc. degree in Computer Science from
considered in excluding or including a paper. University of Duhok, Iraq, in 2008. He completed the M.Sc. degree in Advanced
Computer Science, from Newcastle University, UK, in 2012. He is currently
• Accuracy of data extraction: The individual working as a lecturer at the college of science, University of Duhok, Iraq. His
experience effects extracting the data, therefore main areas of research interest are cloud computing, and software
online meetings were conducted after the data engineering.
Qusay I. Sarhan received the B.Sc. degree in Software Engineering from
extraction process by each author. During the University of Mosul, Iraq, in 2007 and the M.Tech. degree in Software
meetings, the outcomes from each author were Engineering from Jawaharlal Nehru Technological University, India, in 2011.
compared with other findings to determine the Currently, he is a lecturer and the leader of Software Engineering and
Embedded Systems (SEES) research group at University of Duhok, Iraq. He has
differences and reach a final consensus. a couple of national and international publications and his research interests
• Reproducibility of the study: Whether other include software engineering, internet of things, and embedded systems.
researchers could obtain similar outcomes of this
Funding
study is another threat. Thus, to address this, the Not applicable.
research methodology contains the well-explained
steps and actions conducted in this paper (as shown Availability of data and materials
Not applicable.
in “Research methodology” section).

Conclusions Declarations
The contributions of the work presented in this paper
Competing interests
are threefold: (a) a methodical review of related litera- The authors declare that they have no competing interests.
ture on the topic of serverless computing, to address the
issue of the lack of compiling information on the state- Author details
1 Software Engineering and Embedded Systems (SEES) Research Group,
of-the-art of the field; (b) a comparison of the platforms College of Medicine, University of Duhok, Duhok, Kurdistan Region, Iraq.
and tools used in serverless computing; (c) an extensive 2 Software Engineering and Embedded Systems (SEES) Research Group,

analysis of the differences, benefits, and issues related to Department of Computer Science, College of Science, University of Duhok,
Duhok, Kurdistan Region, Iraq.
serverless computing, to provide a more complete under-
standing of the topic. Given the fast evolution and growing Received: 1 July 2020 Accepted: 21 June 2021
interest in the field, this survey focused on gathering the
most outstanding trends and outcomes of serverless com-
puting, as described by recent researchers. This survey References
1. Großmann M, Ioannidis C, Le DT (2019) Applicability of serverless
could significantly reduce ambiguity and the entry barrier computing in fog computing environments for iot scenarios. In:
for novice developers to adapt to the serverless environ- Proceedings of the 12th IEEE/ACM International Conference on Utility
ment. Furthermore, the findings presented in this study and Cloud Computing Companion (UCC ‘19 Companion). Association
for Computing Machinery, New York. pp 29–34. https://doi.org/10.1145/
could be of great value for future researchers interested 3368235.3368834

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 21 of 29

2. Boza EF, Abad CL, Villavicencio M, Quimba S, Plaza JA (2017) Reserved, 18. Feng L, Kudva P, Da Silva D, Hu J (2018) Exploring serverless computing
on demand or serverless: Model-based simulations for cloud budget for neural network training. In: 2018 IEEE 11th International Conference
planning. In: 2017 IEEE Second Ecuador Technical Chapters Meeting on Cloud Computing (CLOUD). pp 334–341. https://doi.org/10.1109/
(ETCM). pp 1–6. https://doi.org/10.1109/ETCM.2017.8247460 CLOUD.2018.00049
3. Villamizar M, Garcés O, Ochoa L, Castro H, Salamanca L, Verano M, 19. Ao L, Izhikevich L, Voelker GM, Porter G (2018) Sprocket: A serverless
Casallas R, Gil S, Valencia C, Zambrano A, Lang M (2017) Cost video processing framework. In: Proceedings of the ACM Symposium on
comparison of running web applications in the cloud using monolithic, Cloud Computing (SoCC ‘18). Association for Computing Machinery,
microservice, and aws lambda architectures. SOCA 11(2):233–247. New York. pp 263–274. https://doi.org/10.1145/3267809.3267815
https://doi.org/10.1007/s11761-017-0208-y 20. Werner S, Kuhlenkamp J, Klems M, Müller J, Tai S (2018) Serverless big
4. Jonas E, Schleier-Smith J, Sreekanti V, Tsai C-C, Khandelwal A, Pu Q, data processing using matrix multiplication as example. In: 2018 IEEE
Shankar V, Carreira J, Krauth K, Yadwadkar N, Gonzalez JE, Popa RA, International Conference on Big Data (Big Data). pp 358–365. https://doi.
Stoica I, Patterson DA (2019) Cloud Programming Simplified: A Berkeley org/10.1109/BigData.2018.8622362
View on Serverless Computing. http://arxiv.org/abs/1902.03383. 21. Al-Ali Z, Goodarzy S, Hunter E, Ha S, Han R, Keller E, Rozner E (2018)
Accessed 6 Jan 2021 Making serverless computing more serverless. In: 2018 IEEE 11th
5. Geng X, Ma Q, Pei Y, Xu Z, Zeng W, Zou J (2018) Research on early warning International Conference on Cloud Computing (CLOUD). pp 456–459.
system of power network overloading under serverless architecture. https://doi.org/10.1109/CLOUD.2018.00064
In: 2018 2nd IEEE Conference on Energy Internet and Energy System 22. Pérez A, Risco S, Naranjo DM, Caballer M, Moltó G (2019) On-premises
Integration (EI2). pp 1–6. https://doi.org/10.1109/EI2.2018.8582355 serverless computing for event-driven data processing applications. In:
6. Kulkarni SG, Liu G, Ramakrishnan KK, Wood T (2019) Living on the edge: 2019 IEEE 12th International Conference on Cloud Computing (CLOUD).
Serverless computing and the cost of failure resiliency. In: 2019 IEEE pp 414–421. https://doi.org/10.1109/CLOUD.2019.00073
International Symposium on Local and Metropolitan Area Networks 23. Glikson A, Nastic S, Dustdar S (2017) Deviceless edge computing:
(LANMAN). pp 1–6. https://doi.org/10.1109/LANMAN.2019.8846970 Extending serverless computing to the edge of the network. In:
7. Baldini I, Castro P, Chang K, Cheng P, Fink S, Ishakian V, Mitchell N, Proceedings of the 10th ACM International Systems and Storage
Muthusamy V, Rabbah R, Slominski A, Suter P (2017) Serverless Conference (SYSTOR ‘17). Association for Computing Machinery, New
Computing: Current Trends and Open Problems. In: Research Advances York. https://doi.org/10.1145/3078468.3078497
in Cloud Computing. Springer, Singapore. pp 1–20. https://doi.org/10. 24. Al-Ameen M, Spillner J (2019) A systematic and open exploration of faas
1007/978-981-10-5026-8_1 research. In: Proceedings of the European Symposium on Serverless
8. Adzic G, Chatley R (2017) Serverless computing: Economic and Computing and Applications (CEUR Workshop Proceedings ; 2330).
architectural impact. In: Proceedings of the 2017 11th Joint Meeting on CEUR-WS, Zurich. pp 30–35. https://doi.org/10.21256/zhaw-3271
Foundations of Software Engineering (ESEC/FSE 2017). Association for 25. Alqaryouti O, Siyam N (2018) Serverless computing and scheduling tasks
Computing Machinery, New York. pp 884–889. https://doi.org/10.1145/ on cloud: A review. Am Sci Res J Eng Technol Sci (ASRJETS) 40(1):235–247
3106237.3117767 26. Taibi D, El Ioini N, Pahl C, Niederkofler J (2020) Patterns for Serverless
9. Jambunathan B, Yoganathan K (2018) Architecture decision on using Functions (Function-as-a-Service): A Multivocal Literature Review. In:
microservices or serverless functions with containers. In: 2018 International Proceedings of the 10th International Conference on Cloud Computing
Conference on Current Trends Towards Converging Technologies and Services Science - Volume 1: CLOSER. pp 181–192. https://doi.org/
(ICCTCT). pp 1–7. https://doi.org/10.1109/ICCTCT.2018.8551035 10.5220/0009578501810192
10. Wolski R, Krintz C, Bakir F, George G, Lin W-T (2019) Cspot: Portable, 27. Hellerstein JM, Faleiro J, Gonzalez JE, Schleier-Smith J, Sreekanti V,
multi-scale functions-as-a-service for iot. In: Proceedings of the 4th Tumanov A, Wu C (2018) Serverless Computing: One Step Forward, Two
ACM/IEEE Symposium on Edge Computing (SEC ‘19). Association for Steps Back. http://arxiv.org/abs/1812.03651. Accessed 4 Oct 2021
Computing Machinery, New York. pp 236–249. https://doi.org/10.1145/ 28. Rajan AP (2020) A review on serverless architectures - function as a
3318216.3363314 service (faas) in cloud computing. Telecommun Comput Electron
11. Yussupov V, Breitenbücher U, Leymann F, Wurster M (2019) A systematic Control 18(1):530–537. https://doi.org/10.12928/telkomnika.v18i1.12169
mapping study on engineering function-as-a-service platforms and 29. Sadaqat M, Colomo-Palacios R, Knudsen LES (2018) Serverless
tools. In: Proceedings of the 12th IEEE/ACM International Conference on Computing: A Multivocal Literature Review. NOKOBIT - Norsk Konferanse
Utility and Cloud Computing (UCC’19). Association for Computing for Organisasjoners Bruk Av Informasjonsteknologi 26(1):1–13
Machinery, New York. pp 229–240. https://doi.org/10.1145/3344341. 30. Scheuner J, Leitner P (2020) Function-as-a-service performance
3368803 evaluation: A multivocal literature review. J Syst Softw 170:110708.
12. Brenner S, Kapitza R (2019) Trust more, serverless. In: Proceedings of the https://doi.org/10.1016/j.jss.2020.110708
12th ACM International Conference on Systems and Storage (SYSTOR 31. Fox GC, Ishakian V, Muthusamy V, Slominski A (2017) Status of Serverless
‘19). Association for Computing Machinery, New York. pp 33–43. https:// Computing and Function-as-a-Service(FaaS) in Industry and Research.
doi.org/10.1145/3319647.3325825 arXiv e-prints:1708–08028. http://arxiv.org/abs/1708.08028. Accessed 6
13. Kuhlenkamp J, Werner S (2018) Benchmarking faas platforms: Call for Jan 2021
community participation. In: 2018 IEEE/ACM International Conference 32. Hassan HB, Barakat SA, Sarhan QI (2021) Serverless Literature Dataset.
on Utility and Cloud Computing Companion (UCC Companion). Zenodo. https://doi.org/10.5281/zenodo.4660553
pp 189–194. https://doi.org/10.1109/UCC-Companion.2018.00055 33. Pedreira O, García F, Brisaboa N, Piattini M (2015) Gamification in software
14. Somma G, Ayimba C, Casari P, Romano SP, Mancuso V (2020) When less engineering - a systematic mapping. Inf Softw Technol 57:157–168
is more: Core-restricted container provisioning for serverless computing. 34. Lopez-herrejon RE, Linsbauer L, Egyed A (2015) A systematic mapping
In: IEEE INFOCOM 2020 - IEEE Conference on Computer study of search-based software engineering for software product lines.
Communications Workshops (INFOCOM WKSHPS). pp 1153–1159. Inf Softw Technol 61:33–51
https://doi.org/10.1109/INFOCOMWKSHPS50562.2020.9162876 35. Bila N, Dettori P, Kanso A, Watanabe Y, Youssef A (2017) Leveraging the
15. Sewak M, Singh S (2018) Winning in the era of serverless computing and serverless architecture for securing linux containers. In: 2017 IEEE 37th
function as a service. In: 2018 3rd International Conference for International Conference on Distributed Computing Systems Workshops
Convergence in Technology (I2CT). pp 1–5. https://doi.org/10.1109/ (ICDCSW). pp 401–404. https://doi.org/10.1109/ICDCSW.2017.66
I2CT.2018.8529465 36. Chang KS, Fink SJ (2017) Visualizing serverless cloud application logs for
16. van Eyk E, Toader L, Talluri S, Versluis L, Uţ A, Iosup A (2018) Serverless is program understanding. In: 2017 IEEE Symposium on Visual Languages
more: From paas to present cloud computing. IEEE Internet Comput and Human-Centric Computing (VL/HCC). pp 261–265. https://doi.org/
22(5):8–17. https://doi.org/10.1109/MIC.2018.053681358 10.1109/VLHCC.2017.8103476
17. Leitner P, Wittern E, Spillner J, Hummer W (2019) A mixed-method 37. Ishakian V, Muthusamy V, Slominski A (2018) Serving deep learning
empirical study of function-as-a-service software development in models in a serverless platform. In: 2018 IEEE International Conference
industrial practice. J Syst Softw 149:340–359. https://doi.org/10.1016/j. on Cloud Engineering (IC2E). pp 257–262. https://doi.org/10.1109/IC2E.
jss.2018.12.013 2018.00052

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 22 of 29

38. Baldini I, Cheng P, Fink SJ, Mitchell N, Muthusamy V, Rabbah R, Suter P, Proceedings of the 2018 USENIX Conference on Usenix Annual Technical
Tardieu O (2017) The serverless trilemma: Function composition for Conference (USENIX ATC ’18). USENIX Association, USA. pp 789–794
serverless computing. In: Proceedings of the 2017 ACM SIGPLAN 54. Wang A, Zhang J, Ma X, Anwar A, Rupprecht L, Skourtis D, Tarasov V, Yan
International Symposium on New Ideas, New Paradigms, and F, Cheng Y (2020) Infinicache: Exploiting ephemeral serverless functions
Reflections on Programming and Software (Onward! 2017). Association to build a cost-effective memory cache. In: 18th USENIX Conference on
for Computing Machinery, New York. pp 89–103. https://doi.org/10. File and Storage Technologies (FAST 20). USENIX Association, Santa
1145/3133850.3133855 Clara. pp 267–281
39. Kanso A, Youssef A (2017) Serverless: Beyond the cloud. In: Proceedings 55. Kuhlenkamp J, Werner S, Borges MC, El Tal K, Tai S (2019) An evaluation
of the 2nd International Workshop on Serverless Computing (WoSC ‘17). of faas platforms as a foundation for serverless big data processing. In:
Association for Computing Machinery, New York. pp 6–10. https://doi. Proceedings of the 12th IEEE/ACM International Conference on Utility
org/10.1145/3154847.3154854 and Cloud Computing (UCC’19). Association for Computing Machinery,
40. Koller R, Williams D (2017) Will serverless end the dominance of linux in New York. pp 1–9. https://doi.org/10.1145/3344341.3368796
the cloud? In: Proceedings of the 16th Workshop on Hot Topics in 56. Werner S, Girke R, Kuhlenkamp J (2020) An evaluation of serverless data
Operating Systems (HotOS ‘17). Association for Computing Machinery, processing frameworks. In: Proceedings of the 2020 Sixth International
New York. pp 169–173. https://doi.org/10.1145/3102980.3103008 Workshop on Serverless Computing (WoSC’20). Association for
41. Mukhi NK, Prabhu S, Slawson B (2017) Using a serverless framework for Computing Machinery, New York. pp 19–24. https://doi.org/10.1145/
implementing a cognitive tutor: Experiences and issues. In: Proceedings 3429880.3430095
of the 2nd International Workshop on Serverless Computing (WoSC ‘17). 57. Kuhlenkamp J, Werner S, Borges MC, Ernst D, Wenzel D (2020)
Association for Computing Machinery, New York. pp 11–15. https://doi. Benchmarking elasticity of faas platforms as a foundation for
org/10.1145/3154847.3154852 objective-driven design of serverless applications. In: Proceedings of the
42. Nadgowda S, Bila N, Isci C (2017) The less server architecture for cloud 35th Annual ACM Symposium on Applied Computing (SAC ’20).
functions. In: Proceedings of the 2nd International Workshop on Association for Computing Machinery, New York. pp 1576–1585. https://
Serverless Computing (WoSC ‘17). Association for Computing Machinery, doi.org/10.1145/3341105.3373948
New York. pp 22–27. https://doi.org/10.1145/3154847.3154850 58. Werner S, Kuhlenkamp J, Pallas F, Anders N, Mucaj N, Tsaplina O, Schmidt
43. Klimovic A, Wang Y, Stuedi P, Trivedi A, Pfefferle J, Kozyrakis C (2018) C, Yildirim K (2020) Diminuendo! tactics in support of faas migrations. In:
Pocket: Elastic ephemeral storage for serverless analytics. In: Proceedings Paasivaara M, Kruchten P (eds). Agile Processes in Software Engineering
of the 12th USENIX Conference on Operating Systems Design and and Extreme Programming – Workshops. Springer, Cham. pp 125–132
Implementation (OSDI’18). USENIX Association, USA. pp 427–444 59. Kuhlenkamp J, Werner S, Tai S (2020) The ifs and buts of less is more: A
44. Yan M, Castro P, Cheng P, Ishakian V (2016) Building a chatbot with serverless computing reality check. In: 2020 IEEE International
serverless computing. In: Proceedings of the 1st International Workshop Conference on Cloud Engineering (IC2E). pp 154–161. https://doi.org/
on Mashups of Things and APIs (MOTA ‘16). Association for Computing 10.1109/IC2E48712.2020.00023
Machinery, New York. pp 1–4. https://doi.org/10.1145/3007203.3007217 60. Pfandzelter T, Bermbach D (2020) tinyfaas: A lightweight faas platform
45. Barcelona-Pons D, García-López P, Ruiz A, Gómez-Gómez A, París G, for edge environments. In: 2020 IEEE International Conference on Fog
Sánchez-Artigas M (2019) Faas orchestration of parallel workloads. In: Computing (ICFC). pp 17–24. https://doi.org/10.1109/ICFC49376.2020.
Proceedings of the 5th International Workshop on Serverless 00011
Computing (WOSC ‘19). Association for Computing Machinery, New 61. Bermbach D, Karakaya A-S, Buchholz S (2020) Using application
York. pp 25–30. https://doi.org/10.1145/3366623.3368137 knowledge to reduce cold starts in faas services. In: Proceedings of the
46. Barcelona-Pons D, Sánchez-Artigas M, París G, Sutra P, García-López P 35th Annual ACM Symposium on Applied Computing (SAC ’20).
(2019) On the faas track: Building stateful distributed applications with Association for Computing Machinery, New York. pp 134–143. https://
serverless architectures. In: Proceedings of the 20th International Middleware doi.org/10.1145/3341105.3373909
Conference (Middleware ‘19). Association for Computing Machinery, 62. Bermbach D, Maghsudi S, Hasenburg J, Pfandzelter T (2020) Towards
New York. pp 41–54. https://doi.org/10.1145/3361525.3361535 auction-based function placement in serverless fog platforms. In: 2020
47. Kaviani N, Kalinin D, Maximilien M (2019) Towards serverless as IEEE International Conference on Fog Computing (ICFC). pp 25–31.
commodity: A case of knative. In: Proceedings of the 5th International https://doi.org/10.1109/ICFC49376.2020.00012
Workshop on Serverless Computing (WOSC ‘19). Association for 63. Garcia Lopez P, Sanchez-Artigas M, Paris G, Barcelona Pons D,
Computing Machinery, New York. pp 13–18. https://doi.org/10.1145/ Ruiz Ollobarren A, Arroyo Pinto D (2018) Comparison of faas
3366623.3368135 orchestration systems. https://doi.org/10.1109/ucc-companion.2018.
48. Byrne A, Nadgowda S, Coskun AK (2020) Ace: Just-in-time serverless 00049
software component discovery through approximate concrete 64. Sampé J, Sánchez-Artigas M, García-López P, París G (2017) Data-driven
execution. In: Proceedings of the 2020 Sixth International Workshop on serverless functions for object storage. In: Proceedings of the 18th
Serverless Computing (WoSC’20). Association for Computing Machinery, ACM/IFIP/USENIX Middleware Conference (Middleware ‘17). Association
New York. pp 37–42. https://doi.org/10.1145/3429880.3430098 for Computing Machinery, New York. pp 121–133. https://doi.org/10.
49. Sánchez-Artigas M, Eizaguirre GT, Vernik G, Stuart L, García-López P 1145/3135974.3135980
(2020) Primula: A Practical Shuffle/Sort Operator for Serverless 65. Sampé J, Vernik G, Sánchez-Artigas M, García-López P (2018) Serverless
Computing. Association for Computing Machinery, New York data analytics in the ibm cloud. In: Proceedings of the 19th International
50. Parás G, Garcáa-López P, Sánchez-Artigas M (2020) Serverless elastic Middleware Conference Industry (Middleware ‘18). Association for
exploration of unbalanced algorithms. In: 2020 IEEE 13th International Computing Machinery, New York. pp 1–8. https://doi.org/10.1145/
Conference on Cloud Computing (CLOUD). pp 149–157. https://doi.org/ 3284028.3284029
10.1109/CLOUD49709.2020.00033 66. Mirabelli ME, García-López P, Vernik G (2020) Bringing scaling
51. López PG, Arjona A, Sampé J, Slominski A, Villard L (2020) Triggerflow: transparency to proteomics applications with serverless computing. In:
Trigger-based orchestration of serverless workflows. In: Proceedings of Proceedings of the 2020 Sixth International Workshop on Serverless
the 14th ACM International Conference on Distributed and Event-Based Computing (WoSC’20). Association for Computing Machinery, New York.
Systems (DEBS ’20). Association for Computing Machinery, New York. pp 55–60. https://doi.org/10.1145/3429880.3430101
pp 3–14. https://doi.org/10.1145/3401025.3401731 67. Carreira J, Fonseca P, Tumanov A, Zhang A, Katz R (2019) Cirrus: A
52. Carver B, Zhang J, Wang A, Anwar A, Wu P, Cheng Y (2020) Wukong: A serverless framework for end-to-end ml workflows. In: Proceedings of
scalable and locality-enhanced framework for serverless parallel the ACM Symposium on Cloud Computing (SoCC ‘19). Association for
computing. In: Proceedings of the 11th ACM Symposium on Cloud Computing Machinery, New York. pp 13–24. https://doi.org/10.1145/
Computing (SoCC ’20). Association for Computing Machinery, New York. 3357223.3362711
pp 1–15. https://doi.org/10.1145/3419111.3421286 68. Sreekanti V, Wu C, Chhatrapati S, Gonzalez JE, Hellerstein JM, Faleiro JM
53. Klimovic A, Wang Y, Kozyrakis C, Stuedi P, Pfefferle J, Trivedi A (2018) (2020) A fault-tolerance shim for serverless computing. In: Proceedings
Understanding ephemeral storage for serverless analytics. In: of the Fifteenth European Conference on Computer Systems (EuroSys

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 23 of 29

’20). Association for Computing Machinery, New York. https://doi.org/10. 82. Cordingly R, Yu H, Hoang V, Sadeghi Z, Foster D, Perez D, Hatchett R,
1145/3342195.3387535 Lloyd W (2020) The serverless application analytics framework: Enabling
69. Ichnowski J, Lee W, Murta V, Paradis S, Alterovitz R, Gonzalez JE, Stoica I, design trade-off evaluation for serverless software. In: Proceedings of the
Goldberg K (2020) Fog robotics algorithms for distributed motion 2020 Sixth International Workshop on Serverless Computing (WoSC’20).
planning using lambda serverless computing. In: 2020 IEEE International Association for Computing Machinery, New York. pp 67–72. https://doi.
Conference on Robotics and Automation (ICRA). pp 4232–4238. https:// org/10.1145/3429880.3430103
doi.org/10.1109/ICRA40945.2020.9196651 83. Toader L, Uta A, Musaafir A, Iosup A (2019) Graphless: Toward serverless
70. Zhang W, Fang V, Panda A, Shenker S (2020) Kappa: A programming graph processing. In: 2019 18th International Symposium on Parallel and
framework for serverless computing. In: Proceedings of the 11th ACM Distributed Computing (ISPDC). pp 66–73. https://doi.org/10.1109/
Symposium on Cloud Computing (SoCC ’20). Association for Computing ISPDC.2019.00012
Machinery, New York. pp 328–343. https://doi.org/10.1145/3419111. 84. van Eyk E, Iosup A, Seif S, Thömmes M (2017) The spec cloud group’s
3421277 research vision on faas and serverless architectures. In: Proceedings of
71. Shankar V, Krauth K, Vodrahalli K, Pu Q, Recht B, Stoica I, Ragan-Kelley J, the 2nd International Workshop on Serverless Computing (WoSC ‘17).
Jonas E, Venkataraman S (2020) Serverless linear algebra. In: Proceedings Association for Computing Machinery, New York. pp 1–4. https://doi.
of the 11th ACM Symposium on Cloud Computing (SoCC ’20). org/10.1145/3154847.3154848
Association for Computing Machinery, New York. pp 281–295. https:// 85. van Eyk E, Iosup A, Abad CL, Grohmann J, Eismann S (2018) A spec rg
doi.org/10.1145/3419111.3421287 cloud group’s vision on the performance challenges of faas cloud
72. Gupta V, Carrano D, Yang Y, Shankar V, Courtade T, Ramchandran K (2020) architectures. In: Companion of the 2018 ACM/SPEC International
Serverless straggler mitigation using error-correcting codes. In: 2020 Conference on Performance Engineering (ICPE ‘18). Association for
IEEE 40th International Conference on Distributed Computing Systems Computing Machinery, New York. pp 21–24. https://doi.org/10.1145/
(ICDCS). pp 135–145. https://doi.org/10.1109/ICDCS47774.2020.00019 3185768.3186308
73. Wu C, Sreekanti V, Hellerstein JM (2020) Transactional causal consistency 86. van Eyk E, Grohmann J, Eismann S, Bauer A, Versluis L, Toader L, Schmitt
for serverless computing. In: Proceedings of the 2020 ACM SIGMOD N, Herbst N, Abad CL, Iosup A (2019) The spec-rg reference architecture
International Conference on Management of Data (SIGMOD ’20). for faas: From microservices and containers to serverless platforms. IEEE
Association for Computing Machinery, New York. pp 83–97. https://doi. Internet Comput 23(6):7–18. https://doi.org/10.1109/MIC.2019.2952061
org/10.1145/3318464.3389710 87. van Eyk E, Scheuner J, Eismann S, Abad CL, Iosup A (2020) Beyond
74. Pu Q, Venkataraman S, Stoica I (2019) Shuffling, fast and slow: Scalable microbenchmarks: The spec-rg vision for a comprehensive serverless
analytics on serverless infrastructure. In: Proceedings of the 16th USENIX benchmark. In: Companion of the ACM/SPEC International Conference
Conference on Networked Systems Design and Implementation on Performance Engineering (ICPE ’20). Association for Computing
(NSDI’19). USENIX Association, USA. pp 193–206 Machinery, New York. pp 26–31. https://doi.org/10.1145/3375555.
75. Lloyd W, Ramesh S, Chinthalapati S, Ly L, Pallickara S (2018) Serverless 3384381
computing: An investigation of factors influencing microservice 88. Eismann S, Grohmann J, van Eyk E, Herbst N, Kounev S (2020) Predicting
performance. In: 2018 IEEE International Conference on Cloud Engineering the costs of serverless workflows. In: Proceedings of the ACM/SPEC
(IC2E). pp 159–169. https://doi.org/10.1109/IC2E.2018.00039 International Conference on Performance Engineering (ICPE ’20).
76. Lloyd W, Vu M, Zhang B, David O, Leavesley G (2018) Improving Association for Computing Machinery, New York. pp 265–276. https://
application migration to serverless computing platforms: Latency doi.org/10.1145/3358960.3379133
mitigation with keep-alive workloads. In: 2018 IEEE/ACM International 89. van Eyk E, Iosup A (2018) Addressing performance challenges in
Conference on Utility and Cloud Computing Companion (UCC serverless computing. In: ICT. OPEN
Companion). pp 195–200. https://doi.org/10.1109/UCC-Companion. 90. Hendrickson S, Sturdevant S, Harter T, Venkataramani V, Arpaci-Dusseau
2018.00056 AC, Arpaci-Dusseau RH (2016) Serverless computation with
77. Al-Masri E, Diabate I, Jain R, Lam MHL, Nathala SR (2018) A serverless iot openlambda. In: 8th USENIX Workshop on Hot Topics in Cloud
architecture for smart waste management systems. In: 2018 IEEE Computing (HotCloud 16). USENIX Association, Denver. pp 1–7
International Conference on Industrial Internet (ICII). pp 179–180. 91. Oakes E, Yang L, Zhou D, Houck K, Harter T, Arpaci-Dusseau AC,
https://doi.org/10.1109/ICII.2018.00034 Arpaci-Dusseau RH (2018) Sock: Rapid task provisioning with
78. Fotouhi M, Chen D, Lloyd WJ (2019) Function-as-a-service application serverless-optimized containers. In: Proceedings of the 2018 USENIX
service composition: Implications for a natural language processing Conference on Usenix Annual Technical Conference (USENIX ATC ‘18).
application. In: Proceedings of the 5th International Workshop on USENIX Association, USA. pp 57–69
Serverless Computing (WOSC ‘19). Association for Computing 92. Oakes E, Yang L, Houck K, Harter T, Arpaci-Dusseau AC, Arpaci-Dusseau
Machinery, New York. pp 49–54. https://doi.org/10.1145/3366623. RH (2017) Pipsqueak: Lean lambdas with large libraries. In: 2017 IEEE 37th
3368141 International Conference on Distributed Computing Systems Workshops
79. Niu X, Kumanov D, Hung L-H, Lloyd W, Yeung KY (2019) Leveraging (ICDCSW). pp 395–400. https://doi.org/10.1109/ICDCSW.2017.32
serverless computing to improve performance for sequence 93. Singhvi A, Khalid J, Akella A, Banerjee S (2020) Snf: Serverless network
comparison. In: Proceedings of the 10th ACM International Conference functions. In: Proceedings of the 11th ACM Symposium on Cloud
on Bioinformatics, Computational Biology and Health Informatics (BCB Computing (SoCC ’20). Association for Computing Machinery, New York.
‘19). Association for Computing Machinery, New York. pp 683–687. pp 296–310. https://doi.org/10.1145/3419111.3421295
https://doi.org/10.1145/3307339.3343465 94. Wang L, Li M, Zhang Y, Ristenpart T, Swift M (2018) Peeking behind the
80. Cordingly R, Yu H, Hoang V, Perez D, Foster D, Sadeghi Z, Hatchett R, curtains of serverless platforms. In: Proceedings of the 2018 USENIX
Lloyd WJ (2020) Implications of programming language selection for Conference on Usenix Annual Technical Conference (USENIX ATC ’18).
serverless data processing pipelines. In: 2020 IEEE Intl Conf on USENIX Association, USA. pp 133–145
Dependable, Autonomic and Secure Computing, Intl Conf on Pervasive 95. Abad CL, Boza EF, van Eyk E (2018) Package-aware scheduling of faas
Intelligence and Computing, Intl Conf on Cloud and Big Data functions. In: Companion of the 2018 ACM/SPEC International
Computing, Intl Conf on Cyber Science and Technology Congress Conference on Performance Engineering (ICPE ‘18). Association for
(DASC/PiCom/CBDCom/CyberSciTech). pp 704–711. https://doi.org/10. Computing Machinery, New York. pp 101–106. https://doi.org/10.1145/
1109/DASC-PICom-CBDCom-CyberSciTech49142.2020.00120 3185768.3186294
81. Cordingly R, Shu W, Lloyd WJ (2020) Predicting performance and cost of 96. Aumala G, Boza E, Ortiz-Avilés L, Totoy G, Abad C (2019) Beyond load
serverless computing functions with saaf. In: 2020 IEEE Intl Conf on balancing: Package-aware scheduling for serverless platforms. In: 2019
Dependable, Autonomic and Secure Computing, Intl Conf on Pervasive 19th IEEE/ACM International Symposium on Cluster, Cloud and Grid
Intelligence and Computing, Intl Conf on Cloud and Big Data Computing (CCGRID). pp 282–291. https://doi.org/10.1109/CCGRID.
Computing, Intl Conf on Cyber Science and Technology Congress 2019.00042
(DASC/PiCom/CBDCom/CyberSciTech). pp 640–649. https://doi.org/10. 97. Alpernas K, Flanagan C, Fouladi S, Ryzhyk L, Sagiv M, Schmitz T, Winstein
1109/DASC-PICom-CBDCom-CyberSciTech49142.2020.00111 K (2018) Secure serverless computing using dynamic information flow

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 24 of 29

control. Proc ACM Program Lang 2(OOPSLA):1–26. https://doi.org/10. 114. Wu M, Mi Z, Xia Y (2020) A survey on serverless computing and its
1145/3276488 implications for jointcloud computing. In: 2020 IEEE International
98. Kaffes K, Yadwadkar NJ, Kozyrakis C (2019) Centralized core-granular Conference on Joint Cloud Computing. pp 94–101. https://doi.org/10.
scheduling for serverless functions. In: Proceedings of the ACM 1109/JCC49151.2020.00023
Symposium on Cloud Computing (SoCC ‘19). Association for Computing 115. Li Z, Chen Q, Xue S, Ma T, Yang Y, Song Z, Guo M (2020) Amoeba:
Machinery, New York. pp 158–164. https://doi.org/10.1145/3357223. Qos-awareness and reduced resource usage of microservices with
3362709 serverless computing. In: 2020 IEEE International Parallel and Distributed
99. Choi S, Shahbaz M, Prabhakar B, Rosenblum M (2019) λ-nic: Interactive Processing Symposium (IPDPS). pp 399–408. https://doi.org/10.1109/
serverless compute on smartnics. In: Proceedings of the ACM SIGCOMM IPDPS47924.2020.00049
2019 Conference Posters and Demos (SIGCOMM Posters and Demos 116. Du D, Yu T, Xia Y, Zang B, Yan G, Qin C, Wu Q, Chen H (2020) Catalyzer:
’19). Association for Computing Machinery, New York. pp 151–152. Sub-millisecond startup for serverless computing with initialization-less
https://doi.org/10.1145/3342280.3342341 booting. In: Proceedings of the Twenty-Fifth International Conference
100. Manner J, Endreß M, Heckel T, Wirtz G (2018) Cold start influencing on Architectural Support for Programming Languages and Operating
factors in function as a service. In: 2018 IEEE/ACM International Systems (ASPLOS ’20). Association for Computing Machinery, New York.
Conference on Utility and Cloud Computing Companion (UCC pp 467–481. https://doi.org/10.1145/3373376.3378512
Companion). pp 181–188. https://doi.org/10.1109/UCC-Companion. 117. Yu T, Liu Q, Du D, Xia Y, Zang B, Lu Z, Yang P, Qin C, Chen H (2020)
2018.00054 Characterizing serverless platforms with serverlessbench. In:
101. Manner J, Kolb S, Wirtz G (2019) Troubleshooting serverless functions: a Proceedings of the 11th ACM Symposium on Cloud Computing (SoCC
combined monitoring and debugging approach. SICS Softw-Intensiv ’20). Association for Computing Machinery, New York. pp 30–44. https://
Cyber-Physical Syst 34(2):99–104. https://doi.org/10.1007/s00450-019- doi.org/10.1145/3419111.3421280
00398-6 118. Liu J, Mi Z, Huang Z, Hua Z, Xia Y (2020) Hcloud: A serverless platform for
102. Winzinger S, Wirtz G (2019) Model-based analysis of serverless jointcloud computing. In: 2020 IEEE International Conference on Joint
applications. In: 2019 IEEE/ACM 11th International Workshop on Cloud Computing. pp 86–93. https://doi.org/10.1109/JCC49151.2020.
Modelling in Software Engineering (MiSE). pp 82–88. https://doi.org/10. 00022
1109/MiSE.2019.00020 119. Pérez A, Moltó G, Caballer M, Calatrava A (2019) A programming model
103. Winzinger S, Wirtz G (2020) Applicability of coverage criteria for and middleware for high throughput serverless computing applications.
serverless applications. In: 2020 IEEE International Conference on Service In: Proceedings of the 34th ACM/SIGAPP Symposium on Applied
Oriented Systems Engineering (SOSE). pp 49–56. https://doi.org/10. Computing (SAC ‘19). Association for Computing Machinery, New York.
1109/SOSE49046.2020.00013 pp 106–113. https://doi.org/10.1145/3297280.3297292
104. Prechtl M, Lichtenthäler R, Wirtz G (2020) Investigating possibilites for 120. Pérez A, Moltó G, Caballer M, Calatrava A (2018) Serverless computing
protecting and hardening installable faas platforms. In: Dustdar S (ed). for container-based architectures. Future Generation Computer Systems
Service-Oriented Computing. Springer, Cham. pp 107–126 83:50–59. https://doi.org/10.1016/j.future.2018.01.022
105. Gias AU, Casale G (2020) Cocoa: Cold start aware capacity planning for 121. Giménez-Alventosa V, Moltó G, Caballer M (2019) A framework and a
function-as-a-service platforms. In: 2020 28th International Symposium performance assessment for serverless mapreduce on aws lambda.
on Modeling, Analysis, and Simulation of Computer and Future Generation Computer Systems 97:259–274. https://doi.org/10.
Telecommunication Systems (MASCOTS). pp 1–8. https://doi.org/10. 1016/j.future.2019.02.057
1109/MASCOTS50786.2020.9285966 122. Naranjo DM, Risco S, de Alfonso C, Pérez A, Blanquer I, Moltó G (2020)
106. Chatley R, Allerton T (2020) Nimbus: Improving the developer Accelerated serverless computing based on gpu virtualization. J Parallel
experience for serverless applications. In: 2020 IEEE/ACM 42nd Distrib Comput 139:32–42. https://doi.org/10.1016/j.jpdc.2020.01.004
International Conference on Software Engineering: Companion 123. Wang H, Niu D, Li B (2019) Distributed machine learning with a
Proceedings (ICSE-Companion). pp 85–88 serverless architecture. In: IEEE INFOCOM 2019 - IEEE Conference on
107. Casale G, Artač M, van den Heuvel W-J, van Hoorn A, Jakovits P, Leymann Computer Communications. pp 1288–1296. https://doi.org/10.1109/
F, Long M, Papanikolaou V, Presenza D, Russo A, Srirama SN, Tamburri INFOCOM.2019.8737391
DA, Wurster M, Zhu L (2020) Radon: rational decomposition and 124. Ghaemi S, Khazaei H, Musilek P (2020) Chainfaas: An open
orchestration for serverless computing. SICS Softw-Intens Cyber-Phys blockchain-based serverless platform. IEEE Access 8:131760–131778.
Syst 35(1):77–87. https://doi.org/10.1007/s00450-019-00413-w https://doi.org/10.1109/ACCESS.2020.3010119
108. Vandebon J, Coutinho JGF, Luk W, Nurvitadhi E, Naik M (2020) Slate: 125. Goli A, Hajihassani O, Khazaei H, Ardakanian O, Rashidi M, Dauphinee T
Managing heterogeneous cloud functions. In: 2020 IEEE 31st (2020) Migrating from monolithic to serverless: A fintech case study. In:
International Conference on Application-specific Systems, Architectures Companion of the ACM/SPEC International Conference on Performance
and Processors (ASAP). pp 141–148. https://doi.org/10.1109/ASAP49362. Engineering (ICPE ’20). Association for Computing Machinery, New York.
2020.00032 pp 20–25. https://doi.org/10.1145/3375555.3384380
109. Kim J, Lee K (2019) Functionbench: A suite of workloads for serverless 126. Mahmoudi N, Khazaei H (2020) Performance modeling of serverless
cloud function service. In: 2019 IEEE 12th International Conference on computing platforms. IEEE Trans Cloud Comput:1–1. https://doi.org/10.
Cloud Computing (CLOUD). pp 502–504. https://doi.org/10.1109/ 1109/TCC.2020.3033373
CLOUD.2019.00091 127. Mahmoudi N, Khazaei H (2020) Temporal performance modelling of
110. Kim J, Park J, Lee K (2019) Network resource isolation in serverless cloud serverless computing platforms. In: Proceedings of the 2020 Sixth
function service. In: 2019 IEEE 4th International Workshops on International Workshop on Serverless Computing (WoSC’20).
Foundations and Applications of Self* Systems (FAS*W). pp 182–187. Association for Computing Machinery, New York. pp 1–6. https://doi.
https://doi.org/10.1109/FAS-W.2019.00051 org/10.1145/3429880.3430092
111. Park J, Kim H, Lee K (2020) Evaluating concurrent executions of multiple 128. Wurster M, Breitenbücher U, Képes K, Leymann F, Yussupov V (2018)
function-as-a-service runtimes with microvm. In: 2020 IEEE 13th Modeling and automated deployment of serverless applications using
International Conference on Cloud Computing (CLOUD). pp 532–536. tosca. In: 2018 IEEE 11th Conference on Service-Oriented Computing
https://doi.org/10.1109/CLOUD49709.2020.00080 and Applications (SOCA). pp 73–80. https://doi.org/10.1109/SOCA.2018.
112. Choi J, Lee K (2020) Evaluation of network file system as a shared data 00017
storage in serverless computing. In: Proceedings of the 2020 Sixth 129. Yussupov V, Breitenbücher U, Hahn M, Leymann F (2019) Serverless
International Workshop on Serverless Computing (WoSC’20). parachutes: Preparing chosen functionalities for exceptional workloads.
Association for Computing Machinery, New York. pp 25–30. https://doi. In: 2019 IEEE 23rd International Enterprise Distributed Object
org/10.1145/3429880.3430096 Computing Conference (EDOC). pp 226–235. https://doi.org/10.1109/
113. Kim J, Lee K (2020) I/o resource isolation of public cloud serverless EDOC.2019.00035
function runtimes for data-intensive applications. Clust Comput 130. Yussupov V, Breitenbücher U, Leymann F, Müller C (2019) Facing the
23(3):2249–2259. https://doi.org/10.1007/s10586-020-03103-4 unplanned migration of serverless applications: A study on portability

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 25 of 29

problems, solutions, and dead ends. In: Proceedings of the 12th 146. Datta P, Kumar P, Morris T, Grace M, Rahmati A, Bates A (2020) Valve:
IEEE/ACM International Conference on Utility and Cloud Computing Securing function workflows on serverless computing platforms. In:
(UCC’19). Association for Computing Machinery, New York. pp 273–283. Proceedings of The Web Conference 2020 (WWW ’20). Association for
https://doi.org/10.1145/3344341.3368813 Computing Machinery, New York. pp 939–950. https://doi.org/10.1145/
131. Spillner J (2019) Serverless computing and cloud function-based 3366423.3380173
applications. In: Proceedings of the 12th IEEE/ACM International 147. Sankaran A, Datta P, Bates A (2020) Workflow integration alleviates
Conference on Utility and Cloud Computing Companion (UCC ‘19 identity and access management in serverless computing. In: Annual
Companion). Association for Computing Machinery, New York. Computer Security Applications Conference (ACSAC ’20). Association for
pp 177–178. https://doi.org/10.1145/3368235.3370269 Computing Machinery, New York. pp 496–509. https://doi.org/10.1145/
132. Murphy S, Persaud L, Martini W, Bosshard B (2020) On the use of web 3427228.3427665
assembly in a serverless context. In: Paasivaara M, Kruchten P (eds). Agile 148. Elgamal T (2018) Costless: Optimizing cost of serverless computing
Processes in Software Engineering and Extreme Programming – through function fusion and placement. In: 2018 IEEE/ACM Symposium
Workshops. Springer, Cham. pp 141–145 on Edge Computing (SEC). pp 300–312. https://doi.org/10.1109/SEC.
133. Spillner J (2020) Resource management for cloud functions with 2018.00029
memory tracing, profiling and autotuning. In: Proceedings of the 2020 149. Dehury CK, Srirama SN, Chhetri TR (2020) Ccodamic: A framework for
Sixth International Workshop on Serverless Computing (WoSC’20). coherent coordination of data migration and computation platforms.
Association for Computing Machinery, New York. pp 13–18. https://doi. Futur Gener Comput Syst 109:1–16. https://doi.org/10.1016/j.future.
org/10.1145/3429880.3430094 2020.03.029
134. Spillner J, Mateos C, Monge DA (2018) Faaster, better, cheaper: The 150. Dehury C, Jakovits P, Srirama SN, Tountopoulos V, Giotis G (2020) Data
prospect of serverless scientific computing and hpc. In: Mocskos E, pipeline architecture for serverless platform. In: Muccini H, Avgeriou P,
Nesmachnow S (eds). High Performance Computing. Springer, Cham. Buhnova B, Camara J, Caporuscio M, Franzago M, Koziolek A, Scandurra
pp 154–168 P, Trubiani C, Weyns D, Zdun U (eds). Software Architecture. Springer,
135. Alder F, Asokan N, Kurnikov A, Paverd A, Steiner M (2019) S-faas: Cham. pp 241–246
Trustworthy and accountable function-as-a-service using intel sgx. In: 151. Sarkar S, Wankar R, Srirama SN, Suryadevara NK (2020) Serverless
Proceedings of the 2019 ACM SIGSAC Conference on Cloud Computing management of sensing systems for fog computing framework. IEEE
Security Workshop (CCSW’19). Association for Computing Machinery, Sensors J 20(3):1564–72. https://doi.org/10.1109/JSEN.2019.2939182
New York. pp 185–199. https://doi.org/10.1145/3338466.3358916 152. Malawski M, Gajek A, Zima A, Balis B, Figiela K (2017) Serverless execution
136. Kuriata A, Illikkal RG (2020) Predictable performance for qos-sensitive, of scientific workflows: Experiments with hyperflow, aws lambda and
scalable, multi-tenant function-as-a-service deployments. In: Paasivaara google cloud functions. Futur Gener Comput Syst:1–13. https://doi.org/
M, Kruchten P (eds). Agile Processes in Software Engineering and 10.1016/j.future.2017.10.029
Extreme Programming – Workshops. Springer, Cham. pp 133–140 153. Moczurad P, Malawski M (2018) Visual-textual framework for serverless
137. Mohan A, Sane H, Doshi K, Edupuganti S, Nayak N, Sukhomlinov V (2019) computation: A luna language approach. In: 2018 IEEE/ACM
Agile cold starts for scalable serverless. In: Proceedings of the 11th International Conference on Utility and Cloud Computing Companion
USENIX Conference on Hot Topics in Cloud Computing (HotCloud’19). (UCC Companion). pp 169–74. https://doi.org/10.1109/UCC-
USENIX Association, USA. p 21 Companion.2018.00052
138. Shahrad M, Fonseca R, Goiri I, Chaudhry G, Batum P, Cooke J, Laureano E, 154. Pawlik M, Banach P, Malawski M (2020) Adaptation of workflow
Tresness C, Russinovich M, Bianchini R (2020) Serverless in the wild: application scheduling algorithm to serverless infrastructure. In:
Characterizing and optimizing the serverless workload at a large cloud Schwardmann U, Boehme C, B. Heras D, Cardellini V, Jeannot E, Salis A,
provider. In: 2020 USENIX Annual Technical Conference (USENIX ATC Schifanella C, Manumachu RR, Schwamborn D, Ricci L, Sangyoon O,
20). pp 205–218 Gruber T, Antonelli L, Scott SL (eds). Euro-Par 2019: Parallel Processing
139. Obetz M, Das A, Castiglia T, Patterson S, Milanova A (2020) Formalizing Workshops. Springer, Cham. pp 345–356
event-driven behavior of serverless applications. In: Brogi A, 155. Akhtar N, Raza A, Ishakian V, Matta I (2020) Cose: Configuring serverless
Zimmermann W, Kritikos K (eds). Service-Oriented and Cloud functions using statistical learning. In: IEEE INFOCOM 2020 - IEEE
Computing. Springer, Cham. pp 19–29 Conference on Computer Communications. pp 129–38. https://doi.org/
140. Das A, Imai S, Patterson S, Wittie MP (2020) Performance optimization for 10.1109/INFOCOM41043.2020.9155363
edge-cloud serverless platforms via dynamic task placement. In: 2020 156. Cadden J, Unger T, Awad Y, Dong H, Krieger O, Appavoo J (2020) Seuss:
20th IEEE/ACM International Symposium on Cluster, Cloud and Internet Skip redundant paths to make serverless fast. In: Proceedings of the
Computing (CCGRID). pp 41–50. https://doi.org/10.1109/CCGrid49817. Fifteenth European Conference on Computer Systems (EuroSys ’20).
2020.00-89 Association for Computing Machinery, New York. https://doi.org/10.
141. Das A, Leaf A, Varela CA, Patterson S (2020) Skedulix: Hybrid cloud 1145/3342195.3392698
scheduling for cost-efficient execution of serverless applications. In: 157. Balla D, Maliosz M, Simon C (2020) Open source faas performance
2020 IEEE 13th International Conference on Cloud Computing (CLOUD). aspects. In: 2020 43rd International Conference on Telecommunications
pp 609–618. https://doi.org/10.1109/CLOUD49709.2020.00090 and Signal Processing (TSP). pp 358–364. https://doi.org/10.1109/
142. Obetz M, Patterson S, Milanova A (2019) Static call graph construction in TSP49548.2020.9163456
aws lambda serverless applications. In: Proceedings of the 11th USENIX 158. Pelle I, Czentye J, Dóka J, Kern A, Gerő BP, Sonkoly B (2020) Operating
Conference on Hot Topics in Cloud Computing (HotCloud’19). USENIX latency sensitive applications on public serverless edge cloud platforms.
Association, USA. p 20 IEEE Internet Things J:1–1. https://doi.org/10.1109/JIOT.2020.3042428
143. Lin W, Krintz C, Wolski R, Zhang M, Cai X, Li T, Xu W (2018) Tracking 159. Balla D, Maliosz M, Simon C, Gehberger D (2020) Tuning runtimes in
causal order in aws lambda applications. In: 2018 IEEE International open source faas. In: Hsu C-H, Kallel S, Lan K-C, Zheng Z (eds). Internet of
Conference on Cloud Engineering (IC2E). pp 50–60. https://doi.org/10. Vehicles. Technologies and Services Toward Smart Cities. Springer,
1109/IC2E.2018.00027 Cham. pp 250–266
144. George G, Bakir F, Wolski R, Krintz C (2020) Nanolambda: Implementing 160. Carver B, Zhang J, Wang A, Cheng Y (2019) In search of a fast and
functions as a service at all resource scales for the internet of things. In: efficient serverless dag engine. In: 2019 IEEE/ACM Fourth International
2020 IEEE/ACM Symposium on Edge Computing (SEC). pp 220–231. Parallel Data Systems Workshop (PDSW). pp 1–10. https://doi.org/10.
https://doi.org/10.1109/SEC50012.2020.00035 1109/PDSW49588.2019.00005
145. Zhang M, Krintz C, Wolski R (2020) Stoic: Serverless teleoperable hybrid 161. Gadepalli PK, McBride S, Peach G, Cherkasova L, Parmer G (2020) Sledge:
cloud for machine learning applications on edge device. In: 2020 IEEE A Serverless-First, Light-Weight Wasm Runtime for the Edge. Association
International Conference on Pervasive Computing and for Computing Machinery, New York
Communications Workshops (PerCom Workshops). pp 1–6. https://doi. 162. Gadepalli PK, Peach G, Cherkasova L, Aitken R, Parmer G (2019)
org/10.1109/PerComWorkshops48775.2020.9156239 Challenges and opportunities for efficient serverless computing at the

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 26 of 29

edge. In: 2019 38th Symposium on Reliable Distributed Systems (SRDS). 179. Lehvä J, Mäkitalo N, Mikkonen T (2018) Case study: Building a serverless
pp 261–2615. https://doi.org/10.1109/SRDS47363.2019.00036 messenger chatbot. In: Garrigós I, Wimmer M (eds). Current Trends in
163. Somu N, Daw N, Bellur U, Kulkarni P (2020) Panopticon: A comprehensive Web Engineering. Springer, Cham. pp 75–86. https://doi.org/10.1007/
benchmarking tool for serverless applications. In: 2020 International 978-3-319-74433-9_6
Conference on COMmunication Systems NETworkS (COMSNETS). 180. Poth A, Schubert N, Riel A (2020) Sustainability efficiency challenges of
pp 144–151. https://doi.org/10.1109/COMSNETS48256.2020.9027346 modern it architectures – a quality model for serverless energy footprint.
164. Bajaj D, Bharti U, Goel A, Gupta SC (2020) Partial migration for In: Yilmaz M, Niemann J, Clarke P, Messnarz R (eds). Systems, Software
re-architecting a cloud native monolithic application into microservices and Services Process Improvement. Springer, Cham. pp 289–301
and faas. In: Badica C, Liatsis P, Kharb L, Chahal D (eds). Information, 181. Deese A (2018) Implementation of unsupervised k-means clustering
Communication and Computing Technology. Springer, Singapore. algorithm within amazon web services lambda. In: 2018 18th IEEE/ACM
pp 111–124 International Symposium on Cluster, Cloud and Grid Computing
165. Daw N, Bellur U, Kulkarni P (2020) Xanadu: Mitigating Cascading Cold (CCGRID). pp 626–632. https://doi.org/10.1109/CCGRID.2018.00093
Starts in Serverless Function Chain Deployments. Association for 182. Jangda A, Pinckney D, Brun Y, Guha A (2019) Formal foundations of
Computing Machinery, New York serverless computing. Proc ACM Program Lang 3(OOPSLA):1–26. https://
166. HoseinyFarahabady M, Lee YC, Zomaya AY, Tari Z (2017) A qos-aware doi.org/10.1145/3360575
resource allocation controller for function as a service (faas) platform. In: 183. Back T, Andrikopoulos V (2018) Using a microbenchmark to compare
Maximilien M, Vallecillo A, Wang J, Oriol M (eds). Service-Oriented function as a service solutions. In: Kritikos K, Plebani P, de Paoli F (eds).
Computing. Springer, Cham. pp 241–255. https://doi.org/10.1007/978- Service-Oriented and Cloud Computing. Springer, Cham. pp 146–160.
3-319-69035-3_17 https://doi.org/10.1007/978-3-319-99819-0_11
167. Kim YK, HoseinyFarahabady MR, Lee YC, Zomaya AY, Jurdak R (2018) 184. Kritikos K, Skrzypek P (2018) A review of serverless frameworks. In: 2018
Dynamic control of cpu usage in a lambda platform. In: 2018 IEEE IEEE/ACM International Conference on Utility and Cloud Computing
International Conference on Cluster Computing (CLUSTER). pp 234–244. Companion (UCC Companion). pp 161–168. https://doi.org/10.1109/
https://doi.org/10.1109/CLUSTER.2018.00041 UCC-Companion.2018.00051
168. Kim YK, HoseinyFarahabady MR, Lee YC, Zomaya AY (2020) Automated 185. Kritikos K, Skrzypek P (2019) Simulation-as-a-service with serverless
fine-grained cpu cap control in serverless computing platform. IEEE computing. In: 2019 IEEE World Congress on Services (SERVICES), vol.
Transactions on Parallel and Distributed Systems 31(10):2289–2301. 2642-939X. pp 200–205. https://doi.org/10.1109/SERVICES.2019.00056
https://doi.org/10.1109/TPDS.2020.2989771 186. Ast M, Gaedke M (2017) Self-contained web components through
169. Suresh A, Somashekar G, Varadarajan A, Kakarla VR, Upadhyay H, Gandhi serverless computing. In: Proceedings of the 2nd International
A (2020) Ensure: Efficient scheduling and autonomous resource Workshop on Serverless Computing (WoSC ‘17). Association for
management in serverless environments. In: 2020 IEEE International Computing Machinery, New York. pp 28–33. https://doi.org/10.1145/
Conference on Autonomic Computing and Self-Organizing Systems 3154847.3154849
(ACSOS). pp 1–10. https://doi.org/10.1109/ACSOS49614.2020.00020 187. Trach B, Oleksenko O, Gregor F, Bhatotia P, Fetzer C (2019) Clemmys:
170. Suresh A, Gandhi A (2019) Fnsched: An efficient scheduler for serverless Towards secure remote execution in faas. In: Proceedings of the 12th
functions. In: Proceedings of the 5th International Workshop on ACM International Conference on Systems and Storage (SYSTOR ‘19).
Serverless Computing (WOSC ’19). Association for Computing Association for Computing Machinery, New York. pp 44–54. https://doi.
Machinery, New York. pp 19–24. https://doi.org/10.1145/3366623. org/10.1145/3319647.3325835
3368136 188. Aditya P, Akkus IE, Beck A, Chen R, Hilt V, Rimac I, Satzke K, Stein M (2019)
171. Hunhoff E, Irshad S, Thurimella V, Tariq A, Rozner E (2020) Proactive Will serverless computing revolutionize nfv? Proc IEEE 107(4):667–678.
serverless function resource management. In: Proceedings of the 2020 https://doi.org/10.1109/JPROC.2019.2898101
Sixth International Workshop on Serverless Computing (WoSC’20). 189. Baresi L, Filgueira Mendonça D, Garriga M (2017) Empowering
Association for Computing Machinery, New York. pp 61–66. https://doi. low-latency applications through a serverless edge computing
org/10.1145/3429880.3430102 architecture. In: De Paoli F, Schulte S, Broch Johnsen E (eds).
172. Tariq A, Pahl A, Nimmagadda S, Rozner E, Lanka S (2020) Sequoia: Service-Oriented and Cloud Computing. Springer, Cham. pp 196–210
Enabling quality-of-service in serverless computing. In: Proceedings of 190. Tang Y, Yang J (2020) Lambdata: Optimizing serverless computing by
the 11th ACM Symposium on Cloud Computing (SoCC ’20). Association making data intents explicit. In: 2020 IEEE 13th International Conference
for Computing Machinery, New York. pp 311–327. https://doi.org/10. on Cloud Computing (CLOUD). pp 294–303. https://doi.org/10.1109/
1145/3419111.3421306 CLOUD49709.2020.00049
173. Crane M, Lin J (2017) An exploration of serverless architectures for 191. Ali A, Pinciroli R, Yan F, Smirni E (2020) Batch: Machine learning inference
information retrieval. In: Proceedings of the ACM SIGIR International serving on serverless platforms with adaptive batching. In: Proceedings
Conference on Theory of Information Retrieval (ICTIR ‘17). Association for of the International Conference for High Performance Computing,
Computing Machinery, New York. pp 241–244. https://doi.org/10.1145/ Networking, Storage and Analysis (SC ’20). pp 1–15
3121050.3121086 192. Christidis A, Moschoyiannis S, Hsu C-H, Davies R (2020) Enabling
174. Kim Y, Lin J (2018) Serverless data analytics with flint. In: 2018 IEEE 11th serverless deployment of large-scale ai workloads. IEEE Access
International Conference on Cloud Computing (CLOUD). pp 451–455. 8:70150–70161. https://doi.org/10.1109/ACCESS.2020.2985282
https://doi.org/10.1109/CLOUD.2018.00063 193. Aske A, Zhao X (2018) Supporting multi-provider serverless computing
175. Król M, Psaras I (2017) Nfaas: Named function as a service. In: on the edge. In: Proceedings of the 47th International Conference on
Proceedings of the 4th ACM Conference on Information-Centric Parallel Processing Companion (ICPP ‘18). Association for Computing
Networking (ICN ‘17). Association for Computing Machinery, New York. Machinery, New York. https://doi.org/10.1145/3229710.3229742
pp 134–144. https://doi.org/10.1145/3125719.3125727 194. Gunasekaran JR, Thinakaran P, Nachiappan NC, Srivatsa Kannan R,
176. Parres-Peredo A, Piza-Davila I, Cervantes F (2019) Building and evaluating Kandemir MT, Das CR (2020) Characterizing bottlenecks in scheduling
user network profiles for cybersecurity using serverless architecture. In: microservices on serverless platforms. In: 2020 IEEE 40th International
2019 42nd International Conference on Telecommunications and Signal Conference on Distributed Computing Systems (ICDCS). pp 1197–1198.
Processing (TSP). pp 164–167. https://doi.org/10.1109/TSP.2019.8768825 https://doi.org/10.1109/ICDCS47774.2020.00195
177. Ivanov V, Smolander K (2018) Implementation of a devops pipeline for 195. Dziurzanski P, Swan J, Indrusiak LS (2018) Value-based manufacturing
serverless applications. In: Kuhrmann M, Schneider K, Pfahl D, Amasaki S, optimisation in serverless clouds for industry 4.0. In: Proceedings of the
Ciolkowski M, Hebig R, Tell P, Klünder J, Küpper S (eds). Product-Focused Genetic and Evolutionary Computation Conference (GECCO ‘18).
Software Process Improvement. Springer, Cham. pp 48–64. https://doi. Association for Computing Machinery, New York. pp 1222–1229. https://
org/10.1007/978-3-030-03673-7_4 doi.org/10.1145/3205455.3205501
178. Chen H, Zhang L-J (2018) Fbaas: Functional blockchain as a service. In: 196. Mujezinović A, Ljubović V (2019) Serverless architecture for workflow
Chen S, Wang H, Zhang L-J (eds). Blockchain – ICBC 2018. Springer, scheduling with unconstrained execution environment. In: 2019 42nd
Cham. pp 243–250. https://doi.org/10.1007/978-3-319-94478-4_17 International Convention on Information and Communication

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 27 of 29

Technology, Electronics and Microelectronics (MIPRO). pp 242–246. on Serverless Computing (WOSC ‘19). Association for Computing
https://doi.org/10.23919/MIPRO.2019.8756833 Machinery, New York. pp 7–12. https://doi.org/10.1145/3366623.3368134
197. Tricomi G, Giosa D, Merlino G, Romeo O, Longo F (2020) Toward a 213. Aytekin A, Johansson M (2019) Exploiting serverless runtimes for
function-as-a-service framework for genomic analysis. In: 2020 IEEE large-scale optimization. In: 2019 IEEE 12th International Conference on
International Conference on Smart Computing (SMARTCOMP). Cloud Computing (CLOUD). pp 499–501. https://doi.org/10.1109/
pp 314–319. https://doi.org/10.1109/SMARTCOMP50058.2020.00070 CLOUD.2019.00090
198. Fingler H, Akshintala A, Rossbach CJ (2019) Usetl: Unikernels for 214. Gabbrielli M, Giallorenzo S, Lanese I, Montesi F, Peressotti M, Zingaro SP
serverless extract transform and load why should you settle for less? In: (2019) No more, no less: A formal model for serverless computing. In:
Proceedings of the 10th ACM SIGOPS Asia-Pacific Workshop on Systems Riis Nielson H, Tuosto E (eds). Coordination Models and Languages.
(APSys ’19). Association for Computing Machinery, New York. pp 23–30. Springer, Cham. pp 148–157. https://doi.org/10.1007/978-3-030-22397-
https://doi.org/10.1145/3343737.3343750 7_9
199. Soltani B, Ghenai A, Zeghib N (2018) Towards distributed containerized 215. Fasogbon P, You Y, Aksu E (2020) 3d human model creation on a
serverless architecture in multi cloud environment. Proc Comput Sci serverless environment. In: 2020 IEEE International Symposium on Mixed
134:121–128. https://doi.org/10.1016/j.procs.2018.07.152. The 15th and Augmented Reality Adjunct (ISMAR-Adjunct). pp 118–122. https://
International Conference on Mobile Systems and Pervasive Computing doi.org/10.1109/ISMAR-Adjunct51615.2020.00044
(MobiSPC 2018) / The 13th International Conference on Future 216. Dash S, Sodhi R, Sodhi B (2020) A serverless cloud computing framework
Networks and Communications (FNC-2018) / Affiliated Workshops for real-time appliance-usage recommendation. In: 2020 21st National
200. Crespo-Cepeda R, Agapito G, Vazquez-Poletti JL, Cannataro M (2019) Power Systems Conference (NPSC). pp 1–6. https://doi.org/10.1109/
Challenges and opportunities of amazon serverless lambda services in NPSC49263.2020.9331847
bioinformatics. In: Proceedings of the 10th ACM International 217. Thong Tran T, Zhang Y-C, Liao W-T, Lin Y-J, Li M-C, Huang H-S (2020) An
Conference on Bioinformatics, Computational Biology and Health autonomous mobile robot system based on serverless computing and
Informatics (BCB ‘19). Association for Computing Machinery, New York. edge computing. In: 2020 21st Asia-Pacific Network Operations and
pp 663–668. https://doi.org/10.1145/3307339.3343462 Management Symposium (APNOMS). pp 334–337. https://doi.org/10.
201. Dash S, Dash DK (2016) Serverless cloud computing framework for smart 23919/APNOMS50412.2020.9236976
grid architecture. In: 2016 IEEE 7th Power India International Conference 218. Witte PA, Louboutin M, Modzelewski H, Jones C, Selvage J, Herrmann FJ
(PIICON). pp 1–6. https://doi.org/10.1109/POWERI.2016.8077240 (2020) An event-driven approach to serverless seismic imaging in the
202. Horovitz S, Amos R, Baruch O, Cohen T, Oyar T, Deri A (2019) Faastest - cloud. IEEE Trans Parallel Distributed Syst 31(9):2032–2049. https://doi.
machine learning based cost and performance faas optimization. In: org/10.1109/TPDS.2020.2982626
Coppola M, Carlini E, D’Agostino D, Altmann J, Bañares JÁ (eds). 219. Ghosh BC, Addya SK, Somy NB, Nath SB, Chakraborty S, Ghosh SK (2020)
Economics of Grids, Clouds, Systems, and Services. Springer, Cham. Caching techniques to improve latency in serverless architectures. In:
pp 171–186. https://doi.org/10.1007/978-3-030-13342-9_15 2020 International Conference on COMmunication Systems NETworkS
203. Bardsley D, Ryan L, Howard J (2018) Serverless performance and (COMSNETS). pp 666–669. https://doi.org/10.1109/COMSNETS48256.
optimization strategies. In: 2018 IEEE International Conference on Smart 2020.9027427
Cloud (SmartCloud). pp 19–26. https://doi.org/10.1109/SmartCloud. 220. Quaresma D, Fireman D, Pereira TE (2020) Controlling garbage
2018.00012 collection and request admission to improve performance of faas
204. Jackson D, Clynch G (2018) An investigation of the impact of language applications. In: 2020 IEEE 32nd International Symposium on Computer
runtime on the performance and cost of serverless functions. In: 2018 Architecture and High Performance Computing (SBAC-PAD).
IEEE/ACM International Conference on Utility and Cloud Computing pp 175–182. https://doi.org/10.1109/SBAC-PAD49847.2020.00033
Companion (UCC Companion). pp 154–160. https://doi.org/10.1109/ 221. Birman Y, Hindi S, Katz G, Shabtai A (2020) Cost-effective malware
UCC-Companion.2018.00050 detection as a service over serverless cloud using deep reinforcement
205. Kaplunovich A (2019) Tolambda–automatic path to serverless learning. In: 2020 20th IEEE/ACM International Symposium on Cluster,
architectures. In: 2019 IEEE/ACM 3rd International Workshop on Cloud and Internet Computing (CCGRID). pp 420–429. https://doi.org/
Refactoring (IWoR). pp 1–8. https://doi.org/10.1109/IWoR.2019.00008 10.1109/CCGrid49817.2020.00-51
206. Lee H, Satyam K, Fox G (2018) Evaluation of production serverless 222. Quang T, Peng Y (2020) Device-driven on-demand deployment of
computing environments. In: 2018 IEEE 11th International Conference serverless computing functions. In: 2020 IEEE International Conference
on Cloud Computing (CLOUD). pp 442–450. https://doi.org/10.1109/ on Pervasive Computing and Communications Workshops (PerCom
CLOUD.2018.00062 Workshops). pp 1–6. https://doi.org/10.1109/PerComWorkshops48775.
207. Asghar T, Rasool S, Iqbal MU, Qayyum Z, Mian AN, Ubakanma G (2018) 2020.9156140
Feasibility of serverless cloud services for disaster management 223. Gunasekaran JR, Thinakaran P, Nachiappan NC, Kandemir MT, Das CR
information systems. In: 2018 IEEE 20th International Conference on High (2020) Fifer: Tackling Resource Underutilization in the Serverless Era.
Performance Computing and Communications; IEEE 16th International Association for Computing Machinery, New York
Conference on Smart City; IEEE 4th International Conference on Data 224. Müller I, Marroquín R, Alonso G (2020) Lambada: Interactive data analytics
Science and Systems (HPCC/SmartCity/DSS). pp 1054–1057. https://doi. on cold data using serverless cloud infrastructure. In: Proceedings of the
org/10.1109/HPCC/SmartCity/DSS.2018.00175 2020 ACM SIGMOD International Conference on Management of Data
208. McGrath G, Brenner PR (2017) Serverless computing: Design, (SIGMOD ’20). Association for Computing Machinery, New York.
implementation, and performance. In: 2017 IEEE 37th International pp 115–130. https://doi.org/10.1145/3318464.3389758
Conference on Distributed Computing Systems Workshops (ICDCSW). 225. Chahal D, Ojha R, Ramesh M, Singhal R (2020) Migrating large deep
pp 405–410. https://doi.org/10.1109/ICDCSW.2017.36 learning models to serverless architecture. In: 2020 IEEE International
209. Christoforou A, Andreou AS (2018) An effective resource management Symposium on Software Reliability Engineering Workshops (ISSREW).
approach in a faas environment. In: ESSCA@UCC. pp 2–8 pp 111–116. https://doi.org/10.1109/ISSREW51248.2020.00047
210. Rajan RAP (2018) Serverless architecture - a revolution in cloud 226. Jain P, Munjal Y, Gera J, Gupta P (2020) Performance analysis of various
computing. In: 2018 Tenth International Conference on Advanced server hosting techniques. Proc Comput Sci 173:70–77. https://doi.org/
Computing (ICoAC). pp 88–93. https://doi.org/10.1109/ICoAC44903. 10.1016/j.procs.2020.06.010. International Conference on Smart
2018.8939081 Sustainable Intelligent Computing and Applications under ICITETM2020
211. Rahman MM, Hasibul Hasan M (2019) Serverless architecture for big data 227. Khatri D, Khatri SK, Mishra D (2020) Potential bottleneck and measuring
analytics. In: 2019 Global Conference for Advancement in Technology performance of serverless computing: A literature study. In: 2020 8th
(GCAT). pp 1–5. https://doi.org/10.1109/GCAT47503.2019.8978443 International Conference on Reliability, Infocom Technologies and
212. Huber F, Körber N, Mock M (2019) Selena: A serverless energy Optimization (Trends and Future Directions) (ICRITO). pp 161–164.
management system. In: Proceedings of the 5th International Workshop https://doi.org/10.1109/ICRITO48877.2020.9197837

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 28 of 29

228. Kehrer S, Zietlow D, Scheffold J, Blochinger W (2020) Self-tuning Cloud Computing (CLOUD). pp 827–830. https://doi.org/10.1109/
serverless task farming using proactive elasticity control. Clust Comput. CLOUD.2018.00113
https://doi.org/10.1007/s10586-020-03158-3 246. Solaiman K, Adnan MA (2020) Wlec: A not so cold architecture to
229. Kelly D, Glavin F, Barrett E (2020) Serverless computing: Behind the mitigate cold start problem in serverless computing. In: 2020 IEEE
scenes of major platforms. In: 2020 IEEE 13th International Conference International Conference on Cloud Engineering (IC2E). pp 144–153.
on Cloud Computing (CLOUD). pp 304–312. https://doi.org/10.1109/ https://doi.org/10.1109/IC2E48712.2020.00022
CLOUD49709.2020.00050 247. Chan A, Wang K-TA, Kumar V (2019) BalloonJVM : Dynamically Resizable
230. Ivan C, Vasile R, Dadarlat V (2019) Serverless computing: An investigation Heap for FaaS. In: CLOUD COMPUTING 2019 : The Tenth International
of deployment environments for web apis. Computers 8(2). https://doi. Conference on Cloud Computing, GRIDs, and Virtualization. pp 99–104
org/10.3390/computers8020050 248. Handoyo E, Arfan M, Soetrisno YAA, Somantri M, Sofwan A, Sinuraya EW
231. Baresi L, Filgueira Mendonça D (2019) Towards a serverless platform for (2018) Ticketing chatbot service using serverless nlp technology. In:
edge computing. In: 2019 IEEE International Conference on Fog 2018 5th International Conference on Information Technology,
Computing (ICFC). pp 1–10. https://doi.org/10.1109/ICFC.2019.00008 Computer, and Electrical Engineering (ICITACEE). pp 325–330. https://
232. Hall A, Ramachandran U (2019) An execution model for serverless doi.org/10.1109/ICITACEE.2018.8576921
functions at the edge. In: Proceedings of the International Conference 249. Zhang M, Zhu Y, Zhang C, Liu J (2019) Video processing with serverless
on Internet of Things Design and Implementation (IoTDI ‘19). computing: A measurement study. In: Proceedings of the 29th ACM
Association for Computing Machinery, New York. pp 225–236. https:// Workshop on Network and Operating Systems Support for Digital Audio
doi.org/10.1145/3302505.3310084 and Video (NOSSDAV ’19). Association for Computing Machinery, New
233. Akkus IE, Chen R, Rimac I, Stein M, Satzke K, Beck A, Aditya P, Hilt V (2018) York. pp 61–66. https://doi.org/10.1145/3304112.3325608
Sand: Towards high-performance serverless computing. In: Proceedings 250. Fan D, He D (2020) A scheduler for serverless framework base on
of the 2018 USENIX Conference on Usenix Annual Technical Conference kubernetes. In: Proceedings of the 2020 4th High Performance
(USENIX ATC ‘18). USENIX Association, USA. pp 923–935 Computing and Cluster Technologies Conference; 2020 3rd
234. Keshavarzian A, Sharifian S, Seyedin S (2019) Modified deep residual International Conference on Big Data and Artificial Intelligence (HPCCT;
network architecture deployed on serverless framework of iot platform BDAI 2020). Association for Computing Machinery, New York.
based on human activity recognition application. Futur Gener Comput pp 229–232. https://doi.org/10.1145/3409501.3409503
Syst 101:14–28. https://doi.org/10.1016/j.future.2019.06.009 251. Thomas S, Ao L, Voelker GM, Porter G (2020) Particle: Ephemeral
235. Palade A, Kazmi A, Clarke S (2019) An evaluation of open source endpoints for serverless networking. In: Proceedings of the 11th ACM
serverless computing frameworks support at the edge. In: 2019 IEEE Symposium on Cloud Computing (SoCC ’20). Association for Computing
World Congress on Services (SERVICES), vol. 2642-939X. pp 206–211. Machinery, New York. pp 16–29. https://doi.org/10.1145/3419111.
https://doi.org/10.1109/SERVICES.2019.00057 3421275
236. Shahrad M, Balkind J, Wentzlaff D (2019) Architectural implications of 252. Gramaglia M, Serrano P, Banchs A, Garcia-Aviles G, Garcia-Saavedra A,
function-as-a-service computing. In: Proceedings of the 52nd Annual Perez R (2020) The case for serverless mobile networking. In: 2020 IFIP
IEEE/ACM International Symposium on Microarchitecture (MICRO ‘52). Networking Conference (Networking). pp 779–784
Association for Computing Machinery, New York. pp 1063–1075. https:// 253. Danayi A, Sharifian S (2018) Pess-mina: A proactive stochastic task
doi.org/10.1145/3352460.3358296 allocation algorithm for faas edge-cloud environments. In: 2018 4th
237. Cicconetti C, Conti M, Passarella A (2020) A decentralized framework for Iranian Conference on Signal Processing and Intelligent Systems
serverless edge computing in the internet of things. IEEE Trans Netw (ICSPIS). pp 27–31. https://doi.org/10.1109/ICSPIS.2018.8700543
Serv Manag:1–1. https://doi.org/10.1109/TNSM.2020.3023305 254. Mendki P (2020) Evaluating webassembly enabled serverless approach
238. De Palma G, Giallorenzo S, Mauro J, Zavattaro G (2020) Allocation priority for edge computing. In: 2020 IEEE Cloud Summit. pp 161–166. https://
policies for serverless function-execution scheduling optimisation. In: doi.org/10.1109/IEEECloudSummit48914.2020.00031
Kafeza E, Benatallah B, Martinelli F, Hacid H, Bouguettaya A, Motahari H 255. Chaudhry SR, Palade A, Kazmi A, Clarke S (2020) Improved qos at the
(eds). Service-Oriented Computing. Springer, Cham. pp 416–430 edge using serverless computing to deploy virtual network functions.
239. Djemame K, Parker M, Datsev D (2020) Open-source serverless IEEE Internet Things J 7(10):10673–10683. https://doi.org/10.1109/JIOT.
architectures: an evaluation of apache openwhisk. In: 2020 IEEE/ACM 2020.3011057
13th International Conference on Utility and Cloud Computing (UCC). 256. Benedict S (2020) Serverless blockchain-enabled architecture for iot
pp 329–335. https://doi.org/10.1109/UCC48980.2020.00052 societal applications. IEEE Trans Comput Soc Syst 7(5):1146–1158.
240. Chadha M, Jindal A, Gerndt M (2020) Towards federated learning using https://doi.org/10.1109/TCSS.2020.3008995
faas fabric. In: Proceedings of the 2020 Sixth International Workshop on 257. Cicconetti C, Conti M, Passarella A (2020) Uncoordinated access to
Serverless Computing (WoSC’20). Association for Computing Machinery, serverless computing in mec systems for iot. Comput Netw 172:107184.
New York. pp 49–54. https://doi.org/10.1145/3429880.3430100 https://doi.org/10.1016/j.comnet.2020.107184
241. Cheng B, Fuerst J, Solmaz G, Sanada T (2019) Fog function: Serverless fog 258. Qiang W, Dong Z, Jin H (2018) Se-lambda: Securing privacy-sensitive
computing for data intensive iot services. In: 2019 IEEE International serverless applications using sgx enclave. In: Beyah R, Chang B, Li Y, Zhu
Conference on Services Computing (SCC). pp 28–35. https://doi.org/10. S (eds). Security and Privacy in Communication Networks. Springer,
1109/SCC.2019.00018 Cham. pp 451–470. https://doi.org/10.1007/978-3-030-01701-9_25
242. Mohanty SK, Premsankar G, di Francesco M (2018) An evaluation of open 259. Prasetyadi G, Hantoro UT, Mutiara AB, Muslim A, Refianti R (2019) Heresy:
source serverless computing frameworks. In: 2018 IEEE International A serverless web application to store compressed and encrypted
Conference on Cloud Computing Technology and Science (CloudCom). document in the form of url. In: 2019 Fourth International Conference
pp 115–120. https://doi.org/10.1109/CloudCom2018.2018.00033 on Informatics and Computing (ICIC). pp 1–5. https://doi.org/10.1109/
243. Li J, Kulkarni SG, Ramakrishnan KK, Li D (2019) Understanding open ICIC47613.2019.8985735
source serverless platforms: Design considerations and performance. In: 260. Kim B, Heo S, Lee J, Jeong S, Lee Y, Kim H (2020) Compiler-assisted
Proceedings of the 5th International Workshop on Serverless semantic-aware encryption for efficient and secure serverless
Computing (WOSC ‘19). Association for Computing Machinery, New computing. IEEE Internet Things J:1–1. https://doi.org/10.1109/JIOT.
York. pp 37–42. https://doi.org/10.1145/3366623.3368139 2020.3031550
244. Nguyen HD, Zhang C, Xiao Z, Chien AA (2019) Real-time serverless: 261. O’Meara W, Lennon RG (2020) Serverless computing security: Protecting
Enabling application performance guarantees. In: Proceedings of the application logic. In: 2020 31st Irish Signals and Systems Conference
5th International Workshop on Serverless Computing (WOSC ’19). (ISSC). pp 1–5. https://doi.org/10.1109/ISSC49989.2020.9180214
Association for Computing Machinery, New York. pp 1–6. https://doi. 262. Agache A, Brooker M, Iordache A, Liguori A, Neugebauer R, Piwonka P,
org/10.1145/3366623.3368133 Popa D-M (2020) Firecracker: Lightweight virtualization for serverless
245. Saha A, Jindal S (2018) Emars: Efficient management and allocation of applications. In: 17th USENIX Symposium on Networked Systems Design
resources in serverless. In: 2018 IEEE 11th International Conference on and Implementation (NSDI 20). USENIX Association, Santa Clara. pp 419–434

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Hassan et al. Journal of Cloud Computing (2021) 10:39 Page 29 of 29

263. Pinto D, Dias JP, Sereno Ferreira H (2018) Dynamic allocation of serverless Conference (GLOBECOM). pp 1–6. https://doi.org/10.1109/
functions in iot environments. https://doi.org/10.1109/euc.2018.00008 GLOBECOM38437.2019.9013156
264. Mejáa A, Marcillo D, Guaño M, Gualotuña T (2020) Serverless based 280. Lynn T, Rosati P, Lejeune A, Emeakaroha V (2017) A preliminary review of
control and monitoring for search and rescue robots. In: 2020 15th enterprise serverless cloud computing (function-as-a-service) platforms.
Iberian Conference on Information Systems and Technologies (CISTI). In: 2017 IEEE International Conference on Cloud Computing Technology
pp 1–6. https://doi.org/10.23919/CISTI49556.2020.9140444 and Science (CloudCom). pp 162–169. https://doi.org/10.1109/
265. Meißner D, Erb B, Kargl F, Tichy M (2018) Retro-λ: An event-sourced CloudCom.2017.15
platform for serverless applications with retroactive computing support. 281. Enes J, Expósito RR, Touriño J (2020) Real-time resource scaling platform
In: Proceedings of the 12th ACM International Conference on for big data workloads on serverless environments. Futur Gener Comput
Distributed and Event-Based Systems (DEBS ‘18). Association for Syst 105:361–379. https://doi.org/10.1016/j.future.2019.11.037
Computing Machinery, New York. pp 76–87. https://doi.org/10.1145/ 282. Elsakhawy M, Bauer M (2020) Faas2f: A framework for defining
3210284.3210285 execution-sla in serverless computing. In: 2020 IEEE Cloud Summit.
266. Zhang T, Xie D, Li F, Stutsman R (2019) Narrowing the gap between pp 58–65. https://doi.org/10.1109/IEEECloudSummit48914.2020.00015
serverless and its state with storage functions. In: Proceedings of the 283. Bhattacharjee A, Chhokra AD, Kang Z, Sun H, Gokhale A, Karsai G (2019)
ACM Symposium on Cloud Computing (SoCC ‘19). Association for Barista: Efficient and scalable serverless serving system for deep learning
Computing Machinery, New York. pp 1–12. https://doi.org/10.1145/ prediction services. In: 2019 IEEE International Conference on Cloud
3357223.3362723 Engineering (IC2E). pp 23–33. https://doi.org/10.1109/IC2E.2019.00-10
267. Grogan J, Mulready C, McDermott J, Urbanavicius M, Yilmaz M, Abgaz Y, 284. Maissen P, Felber P, Kropf P, Schiavoni V (2020) Faasdom: A benchmark
McCarren A, MacMahon ST, Garousi V, Elger P, Clarke P (2020) A suite for serverless computing. In: Proceedings of the 14th ACM
multivocal literature review of function-as-a-service (faas) infrastructures International Conference on Distributed and Event-Based Systems
and implications for software developers. In: Yilmaz M, Niemann J, Clarke (DEBS ’20). Association for Computing Machinery, New York. pp 73–84.
P, Messnarz R (eds). Systems, Software and Services Process https://doi.org/10.1145/3401025.3401738
Improvement. Springer, Cham. pp 58–75 285. Xu Z, Zhang H, Geng X, Wu Q, Ma H (2019) Adaptive function launching
268. Qin S, Wu H, Wu Y, Yan B, Xu Y, Zhang W (2020) Nuka: A generic engine acceleration in serverless computing platforms. In: 2019 IEEE 25th
with millisecond initialization for serverless computing. In: 2020 IEEE International Conference on Parallel and Distributed Systems (ICPADS).
International Conference on Joint Cloud Computing. pp 78–85. https:// pp 9–16. https://doi.org/10.1109/ICPADS47876.2019.00011
doi.org/10.1109/JCC49151.2020.00021 286. Soltani B, Ghenai A, Zeghib N (2018) A migration-based approach to
269. Zuk P, Rzadca K (2020) Scheduling methods to reduce response latency execute long-duration multi-cloud serverless functions. In: Maamri R,
of function as a service. In: 2020 IEEE 32nd International Symposium on Belala F (eds). Proceedings of the 3rd International Conference on
Computer Architecture and High Performance Computing (SBAC-PAD). Advanced Aspects of Software Engineering, ICAASE 2018, Constantine,
pp 132–140. https://doi.org/10.1109/SBAC-PAD49847.2020.00028 Algeria, December 1-2, 2018 (CEUR Workshop Proceedings), vol. 2326.
270. Shen J, Yu H, Zheng Z, Sun C, Xu M, Wang J (2020) Serpens: A pp 42–50
high-performance serverless platform for nfv. In: 2020 IEEE/ACM 28th 287. Martins H, Araujo F, da Cunha PR (2020) Benchmarking serverless
International Symposium on Quality of Service (IWQoS). pp 1–10. computing platforms. J Grid Comput 18(4):691–709. https://doi.org/10.
https://doi.org/10.1109/IWQoS49365.2020.9213030 1007/s10723-020-09523-1
271. Ginzburg S, Freedman MJ (2020) Serverless isn’t server-less: Measuring
and exploiting resource variability on cloud faas platforms. In: Publisher’s Note
Proceedings of the 2020 Sixth International Workshop on Serverless Springer Nature remains neutral with regard to jurisdictional claims in
Computing (WoSC’20). Association for Computing Machinery, New York. published maps and institutional affiliations.
pp 43–48. https://doi.org/10.1145/3429880.3430099
272. Tankov V, Golubev Y, Bryksin T (2019) Kotless: A serverless framework for
kotlin. In: 2019 34th IEEE/ACM International Conference on Automated
Software Engineering (ASE). pp 1110–1113. https://doi.org/10.1109/ASE.
2019.00114
273. Vahidinia P, Farahani B, Aliee FS (2020) Cold start in serverless
computing: Current trends and mitigation strategies. In: 2020
International Conference on Omni-layer Intelligent Systems (COINS).
pp 1–7. https://doi.org/10.1109/COINS49042.2020.9191377
274. Silva P, Fireman D, Pereira TE (2020) Prebaking Functions to Warm the
Serverless Cold Start. Association for Computing Machinery, New York
275. Tan B, Liu H, Rao J, Liao X, Jin H, Zhang Y (2020) Towards lightweight
serverless computing via unikernel as a function. In: 2020 IEEE/ACM 28th
International Symposium on Quality of Service (IWQoS). pp 1–10.
https://doi.org/10.1109/IWQoS49365.2020.9213020
276. Cordasco G, D’Auria M, Negro A, Scarano V, Spagnuolo C (2020) Fly: A
domain-specific language for scientific computing on faas. In:
Schwardmann U, Boehme C, B. Heras D, Cardellini V, Jeannot E, Salis A,
Schifanella C, Manumachu RR, Schwamborn D, Ricci L, Sangyoon O,
Gruber T, Antonelli L, Scott SL (eds). Euro-Par 2019: Parallel Processing
Workshops. Springer, Cham. pp 531–544
277. Nupponen J, Taibi D (2020) Serverless: What it is, what to do and what
not to do. In: 2020 IEEE International Conference on Software
Architecture Companion (ICSA-C). pp 49–50. https://doi.org/10.1109/
ICSA-C50368.2020.00016
278. Reuter A, Back T, Andrikopoulos V (2020) Cost efficiency under mixed
serverless and serverful deployments. In: 2020 46th Euromicro
Conference on Software Engineering and Advanced Applications
(SEAA). pp 242–245. https://doi.org/10.1109/SEAA51224.2020.00049
279. Mahajan K, Figueiredo D, Misra V, Rubenstein D (2019) Optimal pricing
for serverless computing. In: 2019 IEEE Global Communications

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:

1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at

[email protected]