🐛 Logger not logging abrupt TCP Resets on large body requests

[DarthBenro008]

Fiber version 1.14.4 running on Go1.14.6

Issue description
TCP Gets Shutdown abruptly when multipart with relatively huge size file is passed with default Fiber Settings on server side.

Steps to reproduce:

• Run the snippet below
• Send a multipart with relatively huge file as payload

I noticed that when you send a file let's say 20-30mb in multipart from Postman you get an ECONNRESET error instead of the "Entity Too Large" Response .

I was puzzled with this issue as I forgot to increase the body limit in Go Fiber App Settings , however thanks to @renanbastos93 for helping me out in the discord channel.

Real Issue :
On Further Investigation , i found out native fasthttp behaves the same , but informs us in the terminal about the abrupt connection reset which the GoFiber Logger Middleware doesn't . Would be really helpful if this is added in logger to make it easier to debug.

Code snippet

func main() {
	app := fiber.New()

	app.Get("/", func(c *fiber.Ctx) {
		c.Send("Hello, World 👋!")
	})

	app.Post("/api", func(c *fiber.Ctx) {
		file, _ := c.FormFile("file")
		c.SaveFile(file, file.Filename)
	})

	app.Listen(3000)
}

[ReneWerner87]

we could trace the problem back to postman, please search in the existing posts for a solution or open a new one

postmanlabs/postman-app-support#2214
https://github.com/postmanlabs/postman-app-support/issues?q=is%3Aissue+ECONNRESET

with direct curl requests the application will give the response

please use this for logging output

app.Use(middleware.Logger("#${pid} - ${time} ${status} - ${latency} ${method} ${path} ${error}\n"))

---

additionally we found out that the postman problem is time related, if we include a small delay in the output within fasthttp, postman can read the answer correctly