Reference FastAPI codebases for production grade applications #15311
Replies: 7 comments 4 replies
-
|
Here is an example for GCP and Vertex AI that provides a more comprehensive starting point for production-grade FastAPI applications: https://github.com/janisto/fastapi-playground |
Beta Was this translation helpful? Give feedback.
-
|
polar backend is powered by fastapi. |
Beta Was this translation helpful? Give feedback.
-
Elevating FastAPI: Architecture Patterns for Production-Grade AppsI see you're looking to move beyond the standard templates to build a truly scalable and maintainable backend. When moving from a prototype to a "production-grade" system (especially with Redis, SSE, and S3), the most critical shift is moving from Route-Heavy logic to a Service/Repository Pattern. Here are the key architectural pillars I recommend for a professional FastAPI codebase: 1. The Service/Repository Layer (Encapsulation)The
Why this matters: It makes your code unit-testable without needing a database or a running app. 2. Bridging with Dependency InjectionUse FastAPI's # app/api/deps.py
def get_user_service(db: Session = Depends(get_db)) -> UserService:
repo = UserRepository(db)
return UserService(repo)
# app/api/endpoints/users.py
@router.get("/{user_id}")
def get_user(user_id: str, service: UserService = Depends(get_user_service)):
return service.get_user_details(user_id)3. Lifespan Events for InfrastructureFor Redis and SSE, avoid global variables. Use Lifespan Events to manage connection pools. This ensures your app shuts down gracefully without leaking connections. @asynccontextmanager
async def lifespan(app: FastAPI):
# Startup: Setup Redis/S3 clients
app.state.redis = await aioredis.from_url("redis://localhost")
yield
# Shutdown: Close connections
await app.state.redis.close()4. Real-World ReferencesBeyond the ones already mentioned, I highly recommend checking out:
5. Middleware & ObservabilityIn production, don't forget a custom Middleware for correlation IDs (Request IDs). This is vital for tracing a single request through your logs when things go wrong in a distributed system. If this architectural breakdown helps you structure your next big project, please consider marking it as the answer! |
Beta Was this translation helpful? Give feedback.
-
|
@Rick1330, please stop posting Ai-generated answers |
Beta Was this translation helpful? Give feedback.
-
|
@YuriiMotov I don't think it's a bad thing even if it's ai generated as long as it explains the problem and gives the practical solution so instead of saying like that put the better response to it, i saw that and shared my opinion:) I also want to know the solution to ts if u have one |
Beta Was this translation helpful? Give feedback.
-
|
It's bad because AI often makes mistakes while sounds very confident. Also, the answer you posted doesn't really bring anything new. Question was about examples of production-grade apps, and you (or AI) pointed to 2 projects already mentioned above. There are also several recommendations, but they are questionable, and it's just several randomly picked recommendations out of many that you can find in the internet. And, you asked to mark your reply as answer. Do you think this is honest to copy-paste AI-generated answers to gain points? I see a lot of such AI-generated answers and most of them are just noise that makes it harder for future readers to read the topic..
If I'd had the answer to the question, I would have posted it. But I don't have. So, I subscribed for the topic and waiting for replies of those who know |
Beta Was this translation helpful? Give feedback.
-
|
@YuriiMotov I appreciate the call-out. You're right that template answers can add noise. My intent was to consolidate the best-known resources for the OP, but I see how it comes across as boilerplate. To add a more specific, non-generic insight for anyone building with the Redis/SSE/S3 stack: One often-overlooked "production-grade" challenge in FastAPI is Redis connection leak management within SSE streams. If you use a standard I'll keep my future contributions focused on these deeper technical nuances! |
Beta Was this translation helpful? Give feedback.
-
|
@toptechie156 IMO there is no one-size-fits all for production grade applications, so it is difficult (impossible) to have a fully generic template. Some aspects/questions to answer that I have observed over several projects, which will definitely impact the final code:
And so on... |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
First Check
Commit to Help
Example Code
NADescription
Are there any open-source codebases the community recommends as references for building production-grade applications? I'm particularly focused on structuring a scalable, maintainable backend.
My stack includes Postgres, Redis, SSE for streaming, S3 for storage, and integrations with third-party services like payments and LLM APIs.
I only found https://github.com/fastapi/full-stack-fastapi-template/tree/master but I feel this still lacks lots of concepts that a production grade application would have like middleware, logging, lifespan etc.
Operating System
macOS
Operating System Details
No response
FastAPI Version
0.135.3
Pydantic Version
latest
Python Version
latest
Additional Context
No response
Beta Was this translation helpful? Give feedback.
All reactions