Skip to content

architecture-patterns.md

Latest commit

 

History

History
105 lines (81 loc) · 3.93 KB

architecture-patterns.md

File metadata and controls

105 lines (81 loc) · 3.93 KB
Raw

Architecture patterns

  • Patterns of Enterprise Application Architecture ( Martin Fowler )
  • Software Architecture Patterns ( Mark Richards )
  • Enterprise Integration Patterns ( Gregor Hohpe )

Antipatterns

architecture antipatterns

Architecture phases

architecture phases

Architecture styles

  • architecture styles

:TODO: cell-based Architecture (cellular architecture)

Architecture style: Microservice

Microservices Benefits:

  • Independent Deployments
  • Fault Isolation
  • Enchanced Scalability

Microservices importance of design patterns:

  • Scalability
  • Reducing Complexity
  • Distributed Data Management
  • Enhancing Communication

Microservices Design patterns:

  • API Gateway

    like GoF.facade single entry point for all client requests

  • Database per Service

    like a GoF.memento ( partially ) single databank per service, data isolation

  • Circuit Breaker

    prevent overwhelming of the calls to outdated external resource

  • Event-Driven

    like GoF.observer publish event ( notify ), when own state has changed

  • Saga

    like a GoF.command + GoF.proxy for list of the external call will make undo in case of fail

Model Driven Architectre

RM-ODP - Reference Model for Open Distributed Processing

It is an international standard used to design large-scale, distributed systems (systems running on multiple computers or servers across networks).
RM-ODP divides a system into 5 distinct Viewpoints. Instead of looking at the whole chaotic system at once, you look at it through five different "lenses":

Viewpoint What it Focuses On The Core Question
Enterprise Business goals, budget, scope, and organizational policies. Why are we building this, and what business rules apply?
Information The data structure, semantics, and information flow. What information does the system need to store and process?
Computational Functional breakdown and how separate components interact. What do the individual software blocks actually do?
Engineering The underlying infrastructure, communication channels, and deployment. How do we physically move data between servers reliably and securely?
Technology The specific hardware, software, and tech stack choices. Are we using AWS or Azure? Python or Java? Linux or Windows?

Communication between applications

File System

Shared DB

RMI - remote method invocation

REST

can be considered as a RemoteMethodInvocation

flowchart LR

a[component A] --->|triggs| b[component B]
Loading
  • both of the services must be online
  • for scaling need to add additional layer ( facade/InformationExpert )

messaging

Kafka

Model: Publisher(1) --> Subscriber(*)
"Component B" must ask by himself the Broker

flowchart LR

a(component A) --->|triggs| br(broker)

br <---|triggs| b(component B) 

m[message]
a -.->|create| m  -.->|read| b
Loading

local start is cumbersome

JMS ( or even ReDis stream)

Model: Producer(1) --> Consumer(+)

flowchart LR

a(component A) --->|triggs| br(broker)

br --->|triggs| b(component B) 

m[message]
a -.->|create| m  -.->|read| b
Loading