Skip to content
Justin's Digital Garden

Case Study: Visualize Complex Microservice Data Using Python

Source: My personal notes from Case Study: Visualize Complex Microservice Data Using Python Online Class | LinkedIn Learning, formerly Lynda.com

Use Case

Section titled “Use Case”

Visualizing user datto help users or troubleshoot issues with the data.

Create a solution from an idea at work, document it, develop it, and present and improve with others

Why use Python? ease of use and other engineers are likely to be familiar with it; it is good for automation

Using Microservices

Section titled “Using Microservices”

Use case for microservices: Apps than are complex and need to scale, large applications with many functions. It can be hard to understand the data and their relationships. A visualization in a diagram can help.

Microservices vs Monolithic applications

Section titled “Microservices vs Monolithic applications”

Advantages of microservices:

Major functions separated to manage complexity, easier to scale and provide availability for different components

Disadvantages of microservices:

  • Increased network communication between services and security - performance overhead
  • Due to distributed nature of microservice, requires management like through people management, automation, monitoring, and service integration
    • Data is distributed
    • Multiple platforms might be used
  • Can be more difficult to understand product and its flows

Use case: How to Visualize Distributed Data?

Section titled “Use case: How to Visualize Distributed Data?”

Use case: Banned user from a social media platform with 250,000 followers wants to view their activity and manually review their activity. The ban can be done from various places like reports, algorithm detection, or user actions.

Solution: Build a diagnostic tool to visualize data from across different platforms. Needs to be:

  • Maintainable: consider language, libraries, can be used in 1+ years
  • Consider existing team’s skills

Using PlantUML

Section titled “Using PlantUML”
  • PlantUML is an open source tool to help create different types of diagrams like sequence, class, components and others
  • Sequence diagram: show interactions in different services
@startuml


Client -> Server: request data
Server -> Client: response with data
Client -> Server: another request data
Server -> Client: another response with data


@enduml

See more and a web based editor at PlantUML

Visualizing Profile and User management Microservices

Section titled “Visualizing Profile and User management Microservices”

Services to help users manage and review their information:

  • Profile Information
  • Post Management
  • Post Moderation

User’s name is blondiebytes and their profile information. Below UML will render a node about the user, 3 reports on their user activity, and reporter information.

@startuml


object blondiebytes
blondiebytes : name = john
blondiebytes : bio = software developer
blondiebytes : link_in_bio = "linktr.ee/blondiebytes"
blondiebytes : is_reported = true


object blondiebytes_report_one
blondiebytes_report_one : is_manual = false
blondiebytes_report_one : reason = Post used a bad word
blondiebytes_report_one : post_text = You can use ASCII in spring boot


object blondiebytes_report_two
blondiebytes_report_two : is_manual = true
blondiebytes_report_two : reason = spam post
blondiebytes_report_two : reported_by = browniebits
blondiebytes_report_two : post_text = Never use C in the spring


object blondiebytes_report_three
blondiebytes_report_three : is_manual = false
blondiebytes_report_three : reason = represented duplicate content from another account
blondiebytes_report_three : post_text = Learn python first
@enduml


object browniebits
blondiebytes : name = Kate
blondiebytes : bio = software developer
blondiebytes : link_in_bio = "linktr.ee/brownibts"
blondiebytes : is_reported = false
blondiebytes : is_followed_by_blondiebytes = false
blondiebytes : is_following_blondiebytes = true


object browniebits_recent_posts
browniebits_recent_posts : post1 = Spring boot sucks
browniebits_recent_posts : post2 = Never use spring boot


blondiebytes -down-> blondiebytes_report_one
blondiebytes -down-> blondiebytes_report_two
blondiebytes -down-> blondiebytes_report_three


browniebits -down-> browniebits_recent_posts

Using Python with PlantUML

Section titled “Using Python with PlantUML”

Python can be used to generate PlantUML diagrams, see plantuml on pypi.

Python can create the PlantUML code and its diagrams allowing Python developers to use PlantUML more easier. For example, Python can create output PlantUML code as a text file from an input file which can be json.

Doing the Visualization

Section titled “Doing the Visualization”

Using json input data about the user profiles, reports, posts, and their relationships, write a Python script to read the input data and generate a PlantUML node diagram with their relationships. Running the script will read the input data and show the users, their posts, and reports on the user.