David Y.
—How do I use a FastAPI background task within custom FastAPI middleware?
Normally to trigger background tasks in FastAPI, you can simply pass in a BackgroundTasks
object to your function. However, when you are using middleware it’s harder to customize what kind of functions you can pass in. Instead you can create the BackgroundTasks
object manually and assign it to your response by using response.background = background_tasks
. A complete example is as follows:
from fastapi import FastAPI from fastapi import BackgroundTasks from fastapi import Request app = FastAPI() def my_background_task(): print("runs in background") @app.middleware("http") async def my_background_task_middleware(request: Request, call_next): background_tasks = BackgroundTasks() background_tasks.add_task(my_background_task) response = await call_next(request) response.background = background_tasks return response @app.get("/") def hello(): return {"message": "Hello World"}
Below, we describe how background tasks and middleware work in FastAPI in more detail, both separately and in combination.
Sometimes you want to execute some code without delaying the response to the end user. For example, after a user signs up to your web application, you might want to send a “Welcome” email and show a confirmation page to the user. There’s no need to make the user wait until the email has been sent before showing the confirmation page — that can happen in the background.
We can start with a basic “Hello World” application and add a long-running background task. We’ll simulate the task with a sleep(5)
, which will simply wait for five seconds and then print out a status message, but in a real application this code would be doing something useful instead.
from fastapi import FastAPI app = FastAPI() @app.get("/") async def hello(): return {"message": "Hello World"}
A naive way to add the task is simply to call the function before returning a response to the user.
Don’t do this, as it will make the user wait for five seconds unnecessarily.
# Don't do this - the user will have to wait from fastapi import FastAPI from time import sleep app = FastAPI() def my_longrunning_background_function(status: str): sleep(5) print(f"\n\n\n---\nAll done, status {status}\n---\n\n\n") @app.get("/") async def hello(): my_longrunning_background_function("my status") return {"message": "Hello World"}
If you load the page, you’ll see the following things happen:
{"message":"Hello World"}
is shown in the web browser.Instead of the above, you can execute the long-running task in the background without making the user wait:
from fastapi import FastAPI from fastapi import BackgroundTasks from time import sleep app = FastAPI() def my_longrunning_background_function(status: str): sleep(5) print(f"\n\n\n---\nAll done, status {status}\n---\n\n\n") @app.get("/") async def hello(background_tasks: BackgroundTasks): background_tasks.add_task(my_longrunning_background_function, status="my status") return {"message": "Hello World"}
Here we pass a BackgroundTasks
object to our hello()
route. Instead of executing our task directly, it adds it to the background tasks where it will be automatically picked up by FastAPI as soon as possible. Without waiting we return the response to the user.
This will result in the following behavior when you load the page:
{"message": "Hello World"}
in the browser.Middleware is code that automatically gets applied to your main routes without you having to explicitly add it each time. It’s often used for modifying HTTP requests, logging, or any other case where you need to apply the same logic to all (or some category of) existing functions, and functions that you might add in the future.
Here’s an example of a basic custom middleware in FastAPI where we print "I executed"
each time a route is visited:
from fastapi import FastAPI from fastapi import Request app = FastAPI() @app.middleware("http") async def my_middleware(request: Request, call_next): response = await call_next(request) print("I executed") return response @app.get("/") async def hello(): return {"message": "Hello World"} @app.get("/goodbye") async def goodbye(): return {"message": "Goodbye World"}
Note that now we’ve added two routes to our app — hello()
and goodbye()
. We don’t call the middleware function explicitly from either, but if you visit either route you’ll see "I executed"
printed to the logs because the app.middleware("http")
decorator says that we want this middleware to execute for all HTTP requests.
The middleware has access to the Request
object. We generate the response that the calling function would have generated by using call_next(request)
and we can modify or use information from either the request or the response, or execute any other arbitrary code, before returning the final response to the user.
Now that you’ve seen both middleware and background tasks in action, the next example where we combine them will be clearer. Here’s a longer example where we again use sleep
to simulate a background task and have it execute for every route in our application:
from fastapi import FastAPI from fastapi import BackgroundTasks from fastapi import Request from time import sleep app = FastAPI() @app.middleware("http") async def my_longrunning_background_function_middleware(request: Request, call_next): background_tasks = BackgroundTasks() background_tasks.add_task(my_longrunning_background_function, status="my status") response = await call_next(request) response.background = background_tasks return response def my_longrunning_background_function(status: str): sleep(5) print(f"\n\n\n---\nAll done, status {status}\n---\n\n\n") @app.get("/") async def hello(): return {"message": "Hello World"} @app.get("/goodbye") async def goodbye(): return {"message": "Goodbye World"}
As in our first example, we manually create a BackgroundTasks
object within our middleware. We then add our long-running background function to the list of tasks, attach this to the response
object, and then return the response to the user.
For any route the user visits, they will see the response in the browser immediately, but the status message will only appear in the console output five seconds later.
Tasty treats for web developers brought to you by Sentry. Get tips and tricks from Wes Bos and Scott Tolinski.
SEE EPISODESConsidered “not bad” by 4 million developers and more than 100,000 organizations worldwide, Sentry provides code-level observability to many of the world’s best-known companies like Disney, Peloton, Cloudflare, Eventbrite, Slack, Supercell, and Rockstar Games. Each month we process billions of exceptions from the most popular products on the internet.
Here’s a quick look at how Sentry handles your personal information (PII).
×We collect PII about people browsing our website, users of the Sentry service, prospective customers, and people who otherwise interact with us.
What if my PII is included in data sent to Sentry by a Sentry customer (e.g., someone using Sentry to monitor their app)? In this case you have to contact the Sentry customer (e.g., the maker of the app). We do not control the data that is sent to us through the Sentry service for the purposes of application monitoring.
Am I included?We may disclose your PII to the following type of recipients:
You may have the following rights related to your PII:
If you have any questions or concerns about your privacy at Sentry, please email us at [email protected].
If you are a California resident, see our Supplemental notice.