FastAPI seems much slower that jupyter by vscode when useing pymongo

First Check

• I added a very descriptive title to this issue.
• I used the GitHub search to find a similar issue and didn’t find it.
• I searched the FastAPI documentation, with the integrated search.
• I already searched in Google “How to X in FastAPI” and didn’t find any information.
• I already read and followed all the tutorial in the docs and didn’t find an answer.
• I already checked if it is not related to FastAPI but to Pydantic.
• I already checked if it is not related to FastAPI but to Swagger UI.
• I already checked if it is not related to FastAPI but to ReDoc.

Commit to Help

• I commit to help with one of those options 👆

Example Code

@appmat.post("/mt")
def mtest(num: int):
    s = time.perf_counter()
    client = pymongo.MongoClient("mongodb://localhost:27017/")
    db = client['tdm']
    mycollec2 = db['ttt']
    test2Data = []
    for i in range(num):
        test2Data.append({
            "time": i,
            "a": random.randint(1, 8),
            "b": random.randint(0, 1)
        })

    mycollec2.insert_many(test2Data)

    e = time.perf_counter()

    return str(round(sys.getsizeof(test2Data) / 1024 / 1024, 2)) + "MB", e - s

Description

Thanks for help. I am a learner in Python and FastAPI.My work is very simple,using FastAPI to insert data to mongodb by pymongo.

• pymongo version : 4.2.0.
• mongodb version:MongoDB 4.2.8 Community Standalone
• FastAPI Start Command:uvicorn.run('main:app', host='127.0.0.1', port=58000, reload=True)
• CPU:AMD 5800H
• RAM:32G

My problem is that I found inserting data using the way I mentioned above was slow ,but when I copy the code and run in vscode using jupyter directly,it’s much faster than before You can see the detail in the below table:

num	fastapi	jupyter
1000	0.06569519999902695	0.017709599998852354
10000	0.23583729999882053	0.10930559999906109
100000	2.2969444999998814	1.0585518999996566
1000000	23.430175199999212	9.922307700000601
10000000	238.93187810000018	98.82192099999884

So did I do something wrong or how could I make FastAPI running faster when inserting data in my project?

Operating System

Windows

Operating System Details

Window11 22622.598

FastAPI Version

0.85.0

Python Version

Python 3.10.7

Additional Context

No response

[MintBlue]

Sorry for taking everyone’s time,it’s my mistake.I run the code in debugger mode in pycharm. I have not realized that the debugger mode would cause performance lost so much before when using python. And when I running the exe packaged by pyinstall,the cost time is close to jupyter.

Thanks again

[ThirVondukr]

I think in this case there should be no difference between running code in a fastapi endpoint and running the code as-is 🤔