Skip to content

felixvo/event-sourcing-test-app

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Event Sourcing with Go and Redis

NOTE: This code is not tested, just an experiment

I thought you already heard about Event Sourcing in the past recent year. But let's go through the definition again.

Capture all changes to an application state as a sequence of events. Event Sourcing ensures that all changes to application state are stored as a sequence of events. - Martin Fowler

If you know bitcoin/blockchain you will know it's quite similar with Event Sourcing.

Your current balance (Application State) is calculated from a series of events in history (in the chain) Alt Text

so you don't have a table like this in database

user_id balance
10 100$
7 200$

now you have

events
user x top-up event
user buy 5 items event
user y top-up event

I've read many articles/blog posts about Event Sourcing so I try to make once.

What we will build?

Let's say you have an e-commerce website and users can buy items from your website. Source: https://github.com/felixvo/lmax

Entities:

  • User will have balance.
  • Item will have price and number of remain items in the warehouse.

Events:

  • Topup: increase user balance
  • AddItem: add more item to warehouse
  • Order: buy items

Directory Structure

├── cmd
│   ├── consumer       # process events
│   │   ├── handler    # handle new event base on event Type
│   │   └── state
│   └── producer       # publish events
└── pkg
    ├── event          # event definition
    ├── snapshot       # snapshot state of the app
    ├── user           # user domain
    └── warehouse      # item domain

Architecture

Alt Text

Entry IDs The entry ID returned by the XADD command, and identifying univocally >each entry inside a given stream, is composed of two parts: <millisecondsTime>-<sequenceNumber> I use this Entry ID to keep track of processed event

  • The consumer will consume events and build the application state
  • snapshot package will take the application state and save to redis every 30s. Application state will restore from this if our app crash

Run

Producer

First, start the producer to insert some events to redis stream Alt Text

Consumer

Now start the consumer to consume events

Alt Text Because the consumer consumes the events but not backup the state yet. If you wait for more than 30s, you will see this message from console

Alt Text

Now if you stop the app and start it again, the application state will restore from the latest snapshot, not reprocess the event again

Alt Text

Thank you for reading! I hope the source code is clean enough for you to understand 😱

Thoughts

About

sample event sourcing application using go and redis

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Languages