A modern programming language that makes developers happier.

kotlinlang

Hi all!

I have a Ktor server with a Netty engine. On requests, I just call `delay` for some time (say 20 seconds) and then respond with a string.
I thought that when I call `delay`, I can serve other requests while previous requests are suspended, but instead, I have to wait until those requests are done and only then I'm able to serve new requests.
How can I serve new requests when my previous requests are suspended? I thought it works like that by default, but now I really don't understand.

It does work by default. The problem likely lies elsewhere in the server or the client.
You need to share more about your setup to get to the root of this.

Are you calling delay in the middle of a DB transaction so you are blocking accessing to some data in the next call? Just a clue…

```fun main() {
    val req = AtomicInteger(0)
    embeddedServer(Netty, port = 8081, host = "127.0.0.1") {
        routing {
            static("/static") {
                resources()
            }
            get("/wait/{waitMillis}") {
                val r = req.incrementAndGet()
                log.debug("Got request $r")
                val waitMillis = call.parameters["waitMillis"]!!.toLong()
                delay(waitMillis)
                call.respondText("Hello, i've waited for ${waitMillis}ms")
                log.debug("Sent request $r")
            }
        }
    }.start(wait = true)
}```

Here is the code of a test application. When I do a lot of requests, I expect it to finish in `waitMillis` ms, but it processes requests almost one-by-one

Try curl, maybe the browser waits for each

Really, it was a browser issue. Thank you!