Please enable JavaScript.
Coggle requires JavaScript to display documents.
A mixed-method empirical study of Function-as-a-Service software …
A mixed-method empirical study of Function-as-a-Service software
development in industrial practice
Introduction
Cloud
Infrastructure-as-aService (IaaS) / Platform-as-a-Service (PaaS)
“lift-and-shift” style migration
no deep changes
higher level of abstraction
architected
built
deployed
richer development experience
built-in features
“cloud-native”
built for the cloud
serverless computing
entirely transparent
“NoOps”
Function-as-a-Service (FaaS)
AWS Lambda
Background
FaaS characteristics
isolated environment
fine-grained scalability
corresponding pricing
interface
failed executions
automatically retried
events
stateless
Common FaaS system architecture
event controller
events
function container
execution
function datastore
Function source code
user / tenant datastore.
triggers and rules
execution datastore
returned to clients
Study design
characterize the use of FaaS
understanding
Semi-structured interviews
Study protocol
interviews
Analysis
Survey of multi-vocal literature
Article selection
Analysis
Web-based survey
Survey dissemination
Participants
Survey questions
Limitations and threats to validity
External validity
Internal validity
Study results
Mental model
a different mental model
Types of serverless applications
Serverless use case type
Purity
Application patterns
Application size
Common patterns
Development languages and practices
Development challenges
Advantages and challenges
Advantages
Challenges
Deployment costs