Our web app is a FastAPI + React Next.js web app, we deployed it into Railway (https://railway.com/project/aa35baa2-15a4-4062-9b82-2ddd969cc922), but the page response is very slow. We are trying to find out why it is slow.

So we created a mock delay api, it did nothing, only sleep for one second.

Below is the code:

@router.get("/mock-delay", operation_id="mock_delay")

@time_function

async def mock_delay():

"""

Mock API endpoint that simply waits for 1 second and returns.

This endpoint is used to test network latency without any database or processing overhead.

"""

# Sleep for 1 second

await asyncio.sleep(1)

# Return a simple response

return {

"message": "This response was delayed by 1 second",

"status": "success"

}

We also have a decorate in the fastAPI backend to record the time running the api, and it show:

"Function mock_delay completed in decorator 1.000s"

And then we created a very simple page in the React frontend, basically, the page just do

const startTime = performance.now();

await usersApi.mockDelay();

const endTime = performance.now();

then get the Total time: ${(endTime - startTime).toFixed(2)}ms

and the page show:

"Total time: 1894.60ms"

And in your HTTP Logs:

We got:

requestId:

"sM6wgPp6SDCJdvawKxCz9w_3176973899"

timestamp:

"2025-03-06T09:45:27.936970860Z"

method:

"OPTIONS"

path:

"/api/v1/users/mock-delay"

host:

"backend-fastapi-web-production.up.railway.app"

httpStatus:

200

upstreamProto:

"HTTP/1.1"

downstreamProto:

"HTTP/2.0"

responseDetails:

""

totalDuration:

181

upstreamAddress:

"http://[fd12:f94d:27c0:0:2000:40:ba2a:9538]:8000"

clientUa:

"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36"

upstreamRqDuration:

180

txBytes:

2

rxBytes:

694

srcIp:

"49.180.45.204"

edgeRegion:

"asia-southeast1"

and

requestId:

"Oi5E96MdQcGn3NNYHQXw5g_3176973899"

timestamp:

"2025-03-06T09:45:29.358347741Z"

method:

"GET"

path:

"/api/v1/users/mock-delay"

host:

"backend-fastapi-web-production.up.railway.app"

httpStatus:

200

upstreamProto:

"HTTP/1.1"

downstreamProto:

"HTTP/2.0"

responseDetails:

""

totalDuration:

1182

upstreamAddress:

"http://[fd12:f94d:27c0:0:2000:40:ba2a:9538]:8000"

clientUa:

"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36"

upstreamRqDuration:

1182

txBytes:

70

rxBytes:

971

srcIp:

"49.180.45.204"

edgeRegion:

"asia-southeast1"

Our frontend and backend is in the same project (https://railway.com/project/aa35baa2-15a4-4062-9b82-2ddd969cc922). We are using Public Networking for backend and frontend communication.

As the Private Networking never worked for us. Will need your help with this, but we will create another ticket for the private network setting issue.

In this ticket, I was wondering, why our React need 1894ms to finish a 1000ms api call, and why your HTTP logs show 1182ms for GET, and 181ms for OPTIONS.

How much time is spent on the transfer over network. is it normal, how should we optimize and make it faster.