^ @otakusenpai

how do i sleep an app which is using coroutines and ktor for say like 6 secs?

@otakusenpai ^

Or maybe I'm missing something in the docs @Deactivated User? I looked under Structure (advanced..) and Deploy... but no mention of it...

I guess I'll need the full repo around as a reference in the future...

Funny that I have to start a whole application for this though, on one hand it's more flexible, but it means re-installing all the same features for both, etc... it might have been nice to have

routing { port(8080) { ... } ; port(9090) { ...} }

?

I believe you shouldn’t need to start two servers, just provide two connectors. Am I right @cy?

This article on Medium pans ktor, and I wouldn't say it was done fairly: https://medium.freecodecamp.org/going-beyond-android-kotlin-on-the-backend-2a75eef2582b

can someone help me with this one https://hastebin.com/uzetogurup.php ?

Specifically https://github.com/lightbend/config#uses-of-substitutions

And building on that, how does a coroutine know when to suspend and resume?

hey, I've seen people quote this link to say ktor is not performant, but from the data there it seems that they did not even run the tests? https://www.techempower.com/benchmarks/#section=data-r14&hw=ph&test=plaintext

any other benchmarks with comparable results?

is there updated docs or example for new auth api?

Also, if doing blocking processing in an

ApplicationCall

context, what's the better practice, to use suspending functions,

withContext(CommonPool)

, or

async { }

(since some processing can be done in parallel..)?

outdated documentation part https://ktor.io/clients/http-client.html#cio

val client = HttpClient(CIO.config { 
    maxConnectionsCount = 1000 // Maximum number of socket connections.
    endpointConfig = EndpointConfig().apply {
        maxConnectionsPerRoute = 100 // Maximum number of requests for a specific endpoint route.
        pipelineMaxSize = 20 // Max number of opened endpoints.
        keepAliveTime = 5000 // Max number of milliseconds to keep each connection alive.
        connectTimeout = 5000 // Number of milliseconds to wait trying to connect to the server.
        connectRetryAttempts = 5 // Maximum number of attempts for retrying a connection.
    }
})

endpointConfig - Unable to resolve if I understand correctly this line must be replaced to:

endpoint.apply {

can I configure websocket client ping periods and timeouts? websocket client described here https://ktor.io/clients/websockets.html

To use this feature, you need to include io.ktor:ktor-client-websocket artifact.

HttpClient(CIO).get<String>(

I've tried this with both CIO and Apache, I just get an error

No transformation found: class io.ktor.client.engine.apache.ApacheHttpResponse -> class kotlin.String

According to https://ktor.io/clients/http-client.html this should work?

how can i use ssl with raw sockets?

Is there a utility method for creating

ContentType.Application.FormUrlEncoded

strings? I'm currently doing it manually.

Hi there. Is it possible to combine Ktor with Firebase hosting?

Can ktor client limit the number of concurrent connections? Something like this code for HttoAsyncCliebt:

CloseableHttpAsyncClient httpclient = HttpAsyncClients.custom()
    .setMaxConnTotal(100)
    .setMaxConnPerRoute(10)
    .build();

https://hastebin.com/udeyivodop.scala @e5l

Websocket CIO client and netty websocket server - both hides under the hood PING/PONG packets, right? I see this ping pong packets via wireshark but this code never was invoked:

incoming.mapNotNull { it as? Frame.Ping }.consumeEach { frame ->                
                println("ping arrived")
}

I want to use ping/pong packets to determine user online status. it is correct way or no?

And second problem with WebSockets implementation is that pingPeriod and timeout are not interrelated. Maybe I'm wrong... I must be wrong, please)

pingPeriod = Duration.ofSeconds(30) // Disabled (null) by default
timeout = Duration.ofSeconds(15)

When connection between client & server are unexpectedly broken(TCP not closed properly) I expect that an exception will thrown after maximum of 45 seconds.(ping period+*pong* response waiting timeout). But it not happens. What am I do wrong? Should I independently implement this behavior through raw sockets?

Hi guys, I am trying to send a compressed response back to the browser using call.respondWrite like

call.respondWrite(contentType = ContentType.Application.GZip) {
                    val output = WriterOutputStream(this)
                    val writer = BufferedWriter(OutputStreamWriter(GZIPOutputStream(output), "UTF-8"))

...

but this is not compressing the response, do you see anything weird there? or can you point me out to other way of doing it with ktor?

@dave08 The java opts

-XX:+UnlockExperimentalVMOptions -XX:+UseCGroupMemoryLimitForHeap -XX:InitialRAMFraction=2 -XX:MinRAMFraction=2 -XX:MaxRAMFraction=2

can be used on a recent JDK8 as well. As I understand it, JDK10 just makes container awareness automatic.

Here is the OpenJDK issue related to CPU limit support: https://bugs.openjdk.java.net/browse/JDK-8140793

The embedded netty server is very convenient for development (or for someone else to clone the project and quickly run it), but shouldn't be a dependency for production builds. What's the best way to do this? I've been thinking of defining a develop/production property in a gradle settings file, and conditionally including the dependency.