Mock Architecture
- Green button means the database is securely safe and possibly shared-on with another database
- Yellow button means it has several issue based on security, otimization and database trade-off
This is only example to scalable django features based on microservices pattern. Does django with microservices is a good idea? depends on your perspective. Somewhat, django might very bloated or might be overkill (since django its batteries included) when it comes to microservices, another perspective is possible to go with microservices with django if you've got a good reason to do so like you have a large services and you know that you need to scale different components independently or you pretty sure that your database can't handle all that things.
The models example based on Django polls tutorial in django documentation, but i managed it to separated into three services: Polls
, Vote
and Choice
.
Another big different, microservice propose loosely coupled and contradicted against django with tightly coupled style.
Possibly, depending how many traffics or services (that you wanted to scale up) or at least depends on your customization some requirements or features. There's several pros and cons to get going with this pattern, if you got a problem with reusability or customizable - this pattern might come handy to you since you can also independently integrating with any frontend framework or database and also container has a high start-up speed of less than a second. While your data is more important, maybe you can replace it with another pattern like isolated db without docker or shared db
- Docker
- nginx
- MongoDB
- Django REST Framework
Currently, there's a several issue with this project. The big issue is, i don't know how to fetching the data between each service with another services. As Tom Christie said, you can use synchronous request-response in the API, so i didn't know that maybe i'll try it later.
Surprisingly, no. I don't know what will happen if tested and deployed to the production ready, due to many disadvantageous, including in terms of data security, payload between the database and initial integration with client-side.
- Clone this repository
- change directory where
docker-compose.yaml
within folder - Build with
docker-compose build
- After build completed, run
docker-compose up -d
or without detached flagdocker-compose up
- Navigate to localhost, something like
0.0.0.0:8001
forPoll
API and so on