Rodrigo Antonialli presented on microservices architecture. He began by defining microservices as independent services that communicate through lightweight mechanisms like HTTP APIs. Each service focuses on a specific business capability and can be independently deployed. Antonialli then discussed characteristics of microservices like componentization via services, organization around business capabilities, and infrastructure automation. He also covered enabling technologies like containers, messaging systems, and monitoring tools. Finally, Antonialli noted both pros and cons of microservices, such as improved scalability but also increased complexity. He recommended students focus on high-level concepts first and that experienced developers will know how to apply microservices appropriately based on their situation.

1. Introduction to Microservices
Architecture
Rodrigo Antonialli
SECCOMP 2020 - 26-11-2020

2. Agenda
● What is a Microservices Architecture ?
○ Microservices Characteristics
■ Focus on Pros and Cons
■ Enabling Technologies
● Where to go from here

3. Before we begin
● Don’t blindly trust anyone
○ Be responsible for your own choices
● Try to keep your pre-established concepts "under control"
○ You should get to know it, even, and specially, if you don’t like it.
○ Open minded always see something good from it…even if it’s to avoid it.

4. What is a Microservices Architecture?

5. Microservices Architecture
"In short, the microservice architectural style is an approach to developing a single application as a
suite of small services, each running in its own process and communicating with lightweight
mechanisms, often an HTTP resource API. These services are built around business capabilities
and independently deployable by fully automated deployment machinery. There is a bare minimum
of centralized management of these services, which may be written in different programming
languages and use different data storage technologies."
-- James Lewis and Martin Fowler (2014)

6. Microservices Architecture
"Microservices - also known as the microservice architecture - is an architectural style that structures an
application as a collection of services that are
● Highly maintainable and testable
● Loosely coupled
● Independently deployable
● Organized around business capabilities
● Owned by a small team
The microservice architecture enables the rapid, frequent and reliable delivery of large, complex
applications. It also enables an organization to evolve its technology stack. "
-- Chris Richardson - microservices.io

7. Microservices Architecture
● It's not only about how the software components are organized
○ Requires a cultural change
Independent Services
Independent Teams
Projects Products
Highly Coupled Teams
Single Application

8. Microservices Architecture
● Microservices Architecture emerged from the community
○ No single definition, although some consensus
○ Experienced developers who adventured (and had a lot of failures) with SOA
○ Some say: "How to do SOA right"
○ Desired Characteristics and Patterns to do SOA
● Some references with good resources:
○ Martin Fowler & James Lewis - https://martinfowler.com/articles/microservices.html
○ Chris Richardson - https://microservices.io
○ Sam Newman - https://samnewman.io/books/

9. Microservices Architecture
● Most common scenario:
Monolith
Database
Monitoring
Events
Presentation
Business Logic
Data Model
Load Balance
Service A
Database A
Discovery
Service B
Database
Migration

10. Microservices Architecture
● Most common scenario:
Monolith
Presentation
Business Logic
Data Model
Migration
Monolith
Database
Presentation
Business Logic
Data Model
Monolith
Presentation
Business Logic
Data Model
Monitoring
Events
Load Balance
Service A
Database A
Discovery
Service B
Database
Service A
Service A

11. Microservices Architecture
● Where is it adopted today?
○ Big tech companies:
■ Netflix - Shared a lot of tools and libraries they use as open source
■ Uber, Amazon, Ebay, Spotify… and others
○ It's becoming an industry standard solution to solve high scalability and faster delivery
■ There is, however, a lot of discussion about this future
● It’s more suitable for BIG companies with BIG problems to solve
○ It doesn’t mean you can’t start with it - Don’t start with a monolith - Stefan Tilkov
○ You definitely can learn from it

12. Microservices Architecture
● Not a silver bullet - It will not solve all of your problems
● No strict set of rules - No single definition, no single way
● Not the end goal - You should not measure success based on how you “do microservices”
● New set of problems: No receipt. Custom solutions you'll have to figure out by yourself.
○ Patterns established to help with decisions and implementation
Effort and Time
Big organization
Multiple teams
Small organization
Small team(s)
Starting from Scratch
Migrating from a
Monolith

13. Microservices Characteristics

14. Microservices Characteristics - How “micro”?
How “micro” is a microservice?
● 100-200 lines of code?
● Small enough to be re-written in 2 weeks?
● Of course: no single rule to fit it all.
Tips that indicate if two services should be merged:
● Feature redundancy or similarity between features
● High number of calls between services
● Change in interpretation of Business Capabilities and/or Domains
● Constantly updating different services together

15. Microservices Characteristics
● Componentization via Services
● Organization around Business Capabilities
● Products not Projects
● Smart endpoints and dumb pipes
● Decentralized Governance
● Decentralized Data Management
● Infrastructure Automation
● Design for failure
● Evolutionary Design
Microservices - by Martin Fowler

16. Microservices Characteristics - Componentization via Services
Pros
● Highly maintainable and testable
● Independently deployable
● Owned by a small team
● Very flexible scalability
Cons
● Distributed Systems - harder to program, specially with async
communication
● Eventual Consistency - distributed data
● Operational Complexity - large number of services deployed
microservices.io

17. Enabling Technologies
● Containers & Containers Orchestrators
○ Help to automate and simplify deployment - core technologies to enable microservices
○ Docker
○ Podman
○ Docker Compose and Swarm
○ Kubernetes

18. Microservices Characteristics - Organization around Business Capabilities
Pros
● Well-defined scope - favor independent teams work
● Enables faster collaboration - teams can solve design and integrations
faster
Cons
● Might be challenging to define - depends on interpretation
● May slow down at first - the capabilities and domains must be
discussed and defined before multiple teams starting to work
microservices.io

19. Microservices Characteristics - Products not Projects
● Products not Projects
○ Development is not an isolated activity with beginning and end:
■ You build it, you run it
○ The team owns the full life-cycle of the service: design, development, deployment and
monitoring in production
○ Teams also collaborate to product decisions
Pros
● Constant relationship of developers and owned product - easier to
evolve and to provide better user experiments
Cons
● Change in culture - takes time, specially because it involves multiple
teams, developers and managers.

20. Microservices Characteristics - Smart endpoints and dumb pipes
Pros
● Logic and filtering is owned by the service - easier to
understand and change in the future
Cons
● Messaging might pose a different paradigm to build
software - services may follow choreography pattern
instead of orchestration pattern
MS 1
MS 2
MS 3
MS 1
MS 2
MS 3
BUS MS 1
MS 2
MS 3
MS 1
MS 2
MS 3
MS 1
MS 2
MS 3
MS 1
MS 2
MS 3
REST Messages

21. Enabling Technologies
● Messaging
○ Implement event storage, routing and other support functionalities for event driven
communication
■ RabbitMQ
■ Apache Kafka
■ Google Cloud Pub/Sub
■ Amazon SQS

22. Microservices Characteristics - Decentralized Governance
“not every problem is a nail and not every solution a hammer.” - Microservices - Martin Fowler
Pros
● Flexibility to adopt different languages - the right tool for the job
● More opportunities to adopt new technologies
Cons
● Challenges to promote culture of reusage
● Difference between languages might still pose challenges
sometimes
Java Service
Python Service
Go Service
JS Service

23. Enabling Technologies
● Interface Definition Languages - IDL
○ Specification languages used to describe Application Programming Interface (API) - usually
associated to code generation - most common for polyglot environments
■ Google Protocol Buffers
■ Apache Avro
■ Apache Thrift
■ JSON

24. Microservices Characteristics - Decentralized Data Management
Pros
● Flexibility to adopt different models, schemas and storage
technologies - the right tool for the job at data level
Cons
● Distributed Transactions - an operation may involve multiple
services
● Eventual Consistency - needs to be supported/fixed with
compensating operations
● Conceptual alignment may be very hard
Bounded Context - Martin Fowler Microservices - Martin Fowler

25. Microservices Characteristics - Infrastructure Automation
Benefits
● Fast, frequent and reliable delivery - automated tests reduce the
risks of a broken instance to be used - increases the trust on the
deployed software
● Monitoring and tracing automation - faster response to failures
Challenges
● Change in culture to adopt tests as first class citizens (Test
Driven Design)
● Requires knowledge and effort to set up infrastructure beyond
the service domain
Wikipedia - DevOps

26. Enabling Technologies
● Continuous Integration & Continuous Delivery
○ Automate integration and deployment. Heavily used in combination with Containers and
Containers Orchestrations
■ GitHub Actions
■ GitLab Pipelines
■ Jenkins
■ Ansible
■ Puppet

27. Microservices Characteristics - Design for failure
Pros
● Fault tolerance contributes to more responsive application.
● Well-established monitoring helps decrease failures response
time
Cons
● Sync calls - easy to implement, more chance to fail
● Good level of monitoring and distributed tracing may be
challenging to achieve
Circuit Breaker - Martin Fowler

28. Enabling Technologies
● Monitoring & Logging Tools
○ Tools to monitor and debug microservices environments (centralized logging)
○ Grafana
○ Prometheus
○ Zipkin
○ Logstash
○ Graylog

29. Enabling Technologies
● Frameworks
○ Language specific frameworks to make it easy to implement microservices
○ Spring Boot & Spring Cloud - Java
■ Netflix OSS
○ Eclipse Vert.X - Multiple Languages
○ Moleculer - NodeJS
○ What about python? Although you can use Django, Flask or Tornado, to build microservices,
they present themselves as Web frameworks

30. Microservices Characteristics - Evolutionary Design
Pros
● Changes can be more frequent, fast and well-defined.
● Small services can be easily scraped and replaced
● Functionalities can be added as services to existing
monoliths
Cons
● Eventually contracts between services will break
● Versioning - hard for clients when constantly updated - hard for
developer team to support multiple versions
● Monolith to Services - challenging depending on the scenario
microservices.io - Strangler Application Pattern
Service A
Database
Today:
Service B
Database
input
Service A
Database
Tomorrow:
Service B
Database
data
input
Missing required param

31. Microservices Pros
● Flexible Scalability
○ The application doesn’t need to scale homogeneously
○ Scale just what you need
● Improved maintainability
○ Small service is easier to understand and change - makes onboarding easier
● Better testability
○ Faster to test
● Better deployability
○ Services can be deployed independently
○ Reduces breaking changes when best practices of code-review and automated tests are applied.

32. Microservices Pros
● Developers tend to be more productive
○ The IDE is faster, the application starts faster, tests run faster.
○ Less code to keep track and control
● Technology stack is flexible
○ Polyglot Programming: Choose the right tool for the job.
○ Flexible data storage: Choose models and schemas according to context
● Over time, it increases fault tolerance and fault isolation
○ When patterns and best practices are applied and a certain maturity is reached, it’s easier to isolate where/why
a fault happened/is happening.

33. Microservices Cons
● More complex
○ New technologies, new concepts, new patterns - new set of problems to solve
○ Inherits complexities of any distributed systems: latency, failure.
● Requires cultural change
○ Teams might need to be restructured
○ Organization management needs to delegate responsibilities to teams - they own the service, they make
decisions
● More expensive
○ Remote calls - again
○ More resources - independent services, independent CPU, memory, etc.

34. Microservices Cons
● No strict set of rules
○ Still susceptible to different interpretations - disputes on how to adopt/implement
○ It’s hard to choose when/why to adopt
● It may introduce more problems than it solves
○ Maybe your problem is just communication between stakeholders (management and developers)
○ Maybe your problem could be solved just by infrastructure automation

35. Where to go from here

36. Microservices Architecture
Student
Yet another thing to
learn. It’s fine if you
want, but start with
getting the high level
concepts and focus on
the enabling tech.
Junior/Mid-level
You’ll eventually come
across it somewhere,
sometime. It’s helpful
to improve skills and
make choices for your
career.
Experienced
You’ll most likely know
what to do by yourself,
but remember we
never quit learning.
● Should I think / worry / use / learn Microservices Architecture?

37. References
● https://martinfowler.com/articles/microservices.html
● https://microservices.io/
● https://tsh.io/state-of-microservices/
● https://www.tiempodev.com/blog/disadvantages-of-a-microservices-architecture/
● https://dzone.com/articles/30top-tools-for-building-microservices-on-all-leve
● https://medium.com/microservices-architecture/top-10-microservices-framework-for-2020-eefb5e66d
1a2
● https://searchsoftwarequality.techtarget.com/feature/Compare-a-product-vs-project-mindset-for-softw
are-development