Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Dario Abdulrehman
    @dabd
    yep, I read about its excellent package structure
    zfy1355
    @zfy1355
    Hello everyone.i meet a problem today,when i run my app in local,it run good.But when it run in production,it throw a exception like "unexpected message type:PooledUnsafeDirectByteBuf" in customer. Can someone tell me what it is problem?Thanks.
    Jelmer Kuperus
    @jelmerk
    I am stuck with netty version 4.0.33.Final but i'd like to use netty-handler-proxy which does not seem to be available for this old version of netty, is a newer version compatible with this, or am i out of luck
    yeryomenkom
    @yeryomenkom

    Hi!
    Maybe you can help me with one problem...
    My app accumulates a lot of io.netty.util.Recycler.DefaultHandle objects in the heap. And general RAM consumption never decrease until some stop the world GC pause.
    Can you guess am I doing something wrong?

    heap: https://paste.pics/149dfc12708ebfe917297c91e9f6f861
    ram consumption: https://paste.pics/1664079267b8160f5c82b69e35e5ee7c

    yeryomenkom
    @yeryomenkom
    and actually the same goes to io.netty.buffer.PooledUnsafeDirectByteBuf.... I don't understand why do I have so many such objects. I always create them using PooledByteBufAllocator.DEFAULT.directBuffer() and always release them asap. The only thing that comes to my mind is that I create ByteBuf in one thread and after that pass it to another and only after that I pass it to the netty channel...
    Christophe Bornet
    @cbornet

    Hi, I have a small question. If I do

    ByteBuffer buffer = ByteBuffer.allocate(10);
    ByteBuf byteBuf = Unpooled.wrappedBuffer(buffer);
    doSomethingAsync(byteBuf);
    return;

    Is there a possibility of the buffer being GC'd while being used by the async function ?

    Norman Maurer
    @normanmaurer
    @cbornet nope
    Josef Grieb
    @1Jo1
    Hi which cpu architectures are supported for netty on linux? arm and x86?
    Norman Maurer
    @normanmaurer
    aarch64 and x86 for the native transports
    for non native transports whatever is supported by java
    Josef Grieb
    @1Jo1
    does anyone have any idea how to debug java jni? I have tried some gdb java jni tutorial/examples but it didn't work
    Josef Grieb
    @1Jo1
    here is finally my io_uring draft netty/netty#10356 I would appreciate feedback :)
    Francesco Nigro
    @franz1981
    @1Jo1 Try attach to the java process (using the pid of jps) by using CLion!
    kaladhar
    @kaladhar-mummadi
    Can anyone please help me with this ? netty/netty#10350
    Josef Grieb
    @1Jo1
    @franz1981 thanks it worked :)
    Josef Grieb
    @1Jo1
    is there 32 bit support for epoll?
    Willi Schönborn
    @whiskeysierra
    I'm using the LengthFieldPrepender but afterwards I have another custom encoder that adds a leading magic byte and a trailing checksum byte. The problem that I have is that the LengthFieldPrepender emits two events and that produces incorrect messages being written. How can I prevent that?
    Violeta Georgieva
    @violetagg
    Hi,
    I want to support HTTP/2 with TLSv1.2 and TLSv1.3. The specified Http2SecurityUtil.CIPHERS in Netty examples (https://github.com/netty/netty/blob/28e1fbd366602d8f4b50921afb2c9d5e41b94a2d/example/src/main/java/io/netty/example/http2/helloworld/server/Http2Server.java#L60) are TLSv1.2 specific (https://tools.ietf.org/html/rfc7540#section-9.2.2) and not supported by TLSv1.3. As a result when client with TLSv1.3 connects to my server, the handshake fails due to missing TLSv1.3 ciphers. How can I use the JDK TLSv1.3 ciphers in combination with Http2SecurityUtil.CIPHERS?
    javax.net.ssl|DEBUG|1F|nioEventLoopGroup-3-2|2020-07-29 13:46:24.700 EEST|HandshakeContext.java:298|Ignore unsupported cipher suite: TLS_ECDHE_ECDSA_WITH_AES_256_GCM_SHA384 for TLS13
    javax.net.ssl|DEBUG|1F|nioEventLoopGroup-3-2|2020-07-29 13:46:24.700 EEST|HandshakeContext.java:298|Ignore unsupported cipher suite: TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 for TLS13
    javax.net.ssl|DEBUG|1F|nioEventLoopGroup-3-2|2020-07-29 13:46:24.700 EEST|HandshakeContext.java:298|Ignore unsupported cipher suite: TLS_ECDHE_ECDSA_WITH_CHACHA20_POLY1305_SHA256 for TLS13
    javax.net.ssl|DEBUG|1F|nioEventLoopGroup-3-2|2020-07-29 13:46:24.700 EEST|HandshakeContext.java:298|Ignore unsupported cipher suite: TLS_ECDHE_RSA_WITH_CHACHA20_POLY1305_SHA256 for TLS13
    javax.net.ssl|DEBUG|1F|nioEventLoopGroup-3-2|2020-07-29 13:46:24.700 EEST|HandshakeContext.java:298|Ignore unsupported cipher suite: TLS_ECDHE_ECDSA_WITH_AES_128_GCM_SHA256 for TLS13
    javax.net.ssl|DEBUG|1F|nioEventLoopGroup-3-2|2020-07-29 13:46:24.700 EEST|HandshakeContext.java:298|Ignore unsupported cipher suite: TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 for TLS13
    javax.net.ssl|DEBUG|1F|nioEventLoopGroup-3-2|2020-07-29 13:46:24.700 EEST|HandshakeContext.java:305|No available cipher suite for TLS13
    Violeta Georgieva
    @violetagg
    @normanmaurer Can you help with the above ^^^ ?
    ruslan
    @unoexperto
    Hi everyone! I'm using http client from reactor-netty and I wonder about one edge case. So I'm making http request for a large file. In the middle of the request network interface is lost. I expect to get socket exception but instead it's stuck forever. I see that I'm waiting on blocking parameterless select() call in NioEventLoop.select(). I can of course cancel the request from some watcher thread but I wonder if it can be handled in lower level. Could you please advise?
    Prashant Singh
    @prashreyant
    Hey I'm Prashant Singh
    {Ritualistic Continuous Programmer}{Monomaniacal Mad} {Quick Learner} {Idiosyncratic} {Tenacious} {Forthright} {Leadership Quality} {prashreyant}
    I 'm a student of R.E.I Intermediate College
    My Hobbies are: Programming, Brain Storming, Gaming, Travelling....
    My Handles:
    Linkedin :- https://www.linkedin.com/in/prashant-singh-b9a9001b1
    Instagram:- https://www.instagram.com/prashreyant/
    Harsh Maheshwari
    @harshmah
    Hi Folks, My team is using Netty webserver (Along with Spring WebFlux) for the current project. I am trying to understand how this work. Is there any diagram which explains me about its internal component? I have the following question related to this:
    1. How many requests a Netty Web server can accept? Can we cap on this?
    Harsh Maheshwari
    @harshmah
    Any help for this ^^
    francis lee
    @francisoliverlee
    you can test it
    Andrey Mizurov
    @amizurov

    Any help for this ^^

    Hi try asking you question in https://gitter.im/reactor/reactor-netty, because WebFlux not use Netty directly.

    grober82
    @grober82
    unsafe.close(unsafe.voidPromise());
    Atul Gupta
    @atulmgupta
    Hello
    i am trying to set the MDC context to netty pipeline
    can anyone please help
    Andrey Mizurov
    @amizurov
    @atulmgupta could you please share more info
    Artur
    @S0PEX

    Hey it seems that there isn't that much but I am currently working on a basic Chat Client.
    This been said I am unsure about the architecture.
    As I am planning to have multiple chat rooms.
    So the flow would be like this:

    • User creates chat room with a certain name
    • Other users can search for this name and see the chatroom
    • Users can join the chatroom and will receive messages posted in it.

    My idea was to create a ChatRoom object with a ChannelGroup in it.
    When the user enters the chatroom his channel will be added to the channel group inside this ChatRoom.
    When a user sends a message the message will contain a charroom_id so that I can forward it to the correct room.

    As the user can only connect to one chatroom at a time I was wondering if there would be a nice way of automaticly forwarding the messages to the correct room ?

    Currently I am doing a lookup and then invoke the message handler in the correct chatroom. But I am not sure if I should create an own set in the pipeline for this or what the best practise way would be.

    Regards Artur

    gianx89
    @gianx89
    Hi, quick question. I'm using Spring Boot Web Client with Netty. I'm doing a multipart post with 2 text fields and one binary field. The call works with RestTemplate and Postman. I noted that, when I use postman I have:
    image.png
    when I use netty
    image.png
    what does it mean "continuation". Could this be the issue?
    Andrey Mizurov
    @amizurov
    @S0PEX Hi, please take a look at this example https://github.com/netty/netty/tree/4.1/example/src/main/java/io/netty/example/stomp/websocket (STOM + WebSocket chat), maybe you will find something useful.
    Kevin
    @KevinRoebert
    Hi I've a question. When to use ByteBuf.discardReadBytes()? Does this release the memory of the already read part of the underlying byte array?
    Peter
    @warmuuh

    Hi. before netty 1.0.0, we used tcpServer::bootstrap to register a gauge for pending task queue of the eventloop

        factory.addServerCustomizers(httpServer -> httpServer.tcpConfiguration(
            tcpServer -> tcpServer.bootstrap(serverBootstrap -> {
                  serverBootstrap.config().childGroup().forEach(eventExecutor -> {
                    if (SingleThreadEventExecutor.class.isAssignableFrom(eventExecutor.getClass())) {
                      SingleThreadEventExecutor singleThreadEventExecutor = (SingleThreadEventExecutor) eventExecutor;
    
                      Gauge.builder(SERVER_PENDING_TASK_METRIC, singleThreadEventExecutor::pendingTasks)
                          .description("Pending Tasks")
                          .tag(SERVER_THREAD_NAME, singleThreadEventExecutor.threadProperties().name())
                          .tag(SERVER_THREAD_STATE, singleThreadEventExecutor.threadProperties().state().name())
                          .register(registry);
                    }
                  });
                  return serverBootstrap;
                }
            )));

    starting with netty 1.0.0, bootstrap is not available anymore. how can this be done now? thx for help

    2 replies
    Andrey Mizurov
    @amizurov

    Hi I've a question. When to use ByteBuf.discardReadBytes()? Does this release the memory of the already read part of the underlying byte array?

    It depends on implementation, e.g UnpooledHeapByteBuf has capacity 10, readerIndex set to 4 position and writerIndex to 8, so after discardReadBytes() all bytes between readerIndex and writeIndex will be copy to the beginning and readerIndex set to 0 and writerIndex to 4 position.

    Kang Breder
    @kangbreder
    Hello Everyone. My name is Kang Breder Mbulle. I am a student from the University of Buea studying for a M.Eng in Network Engineering and Security. I have saw this organisation on the GSoC page and found it interesting. I will like to contribute to this project as I look forward to GSoC 2021. I have some knowledge of Java and I willing to learn. Can someone please guide me on how to get started?
    Vaibhav Patil
    @vaibhavpatil123
    Hello Colleagous, how we can share common SimpleChannelInboundHandler for more then one client request , at last how we can map request -to-response? in netty
    SimpleChannelInboundHandler
    CHEVALIER
    @et00448
    Hello Everyone
    How Can we desactivate reissue on http 408 response with Netty reactor 0.9.13 ?
    1 reply
    Thank'you
    CHEVALIER
    @et00448
    Hi all,
    Is it possible to accept non URLENCODED with NETTY (like TOMCAT's relaxed-query-chars parameter) ?
    Thanks
    bruto1
    @bruto1
    Hi all,
    any tips on whether or not SslProvider.OPENSSL_REFCNT is production-ready?
    looking to reduce memory and CPU impact of JDK's impl on our service - memory profiling shows it allocating alot of byte[], among other things
    Ilya
    @squadgazzz
    Hello! Is it possible to enable TCP SACK (Selective Acknowledgments) via Netty?
    Marco Zanghì
    @marcozanghi
    Hi all, I am trying to resolve a specific problem with a reactor.netty.http.client.PrematureCloseException when I am using to call an endpoint from a webclient. This endpoint is behind a reverse proxy. Do you have any idea?