FilterChain instantiation

Hi Guys, I am trying to find out some information on how instances of the filter chain is created. 1. Is the filter chain assembled on the acceptor global or is a duplicate created for every session? 2. Does mina wait for an operation on a given session to complete before invoking the next

Re: FilterChain instantiation

I sent this email to the list yesterday but apparently I wasn't fully subscribed. Can someone please forward any responses. Thank you. -- Jon V Sent with Sparrow (http://www.sparrowmailapp.com/?sig) On Tuesday, December 11, 2012 at 9:58 PM, Jon V wrote: Hi Guys, I am trying to find

Re: FilterChain instantiation

is single threaded, do I need to make it aware of multi threaded attempts to invoke it? -- Jon V Sent with Sparrow (http://www.sparrowmailapp.com/?sig) On Wednesday, December 12, 2012 at 6:06 PM, Jon V wrote: I sent this email to the list yesterday but apparently I wasn't fully subscribed

Mina Core 2.0.7 build fails from source zip

I am trying to patch a bug in the IoBuffer caching and it won't build unless I remove the tests from the pom.xml [INFO] [INFO] BUILD FAILURE [INFO]

Patch for IoBuffer management in IoProcessor 2.0.7

Sorry that this is the entire file. I don't have it checked out with subversion to create typical patches. Basically, there are two places where I call free() to release the IoBuffer memory. The CachedBufferAllocator is totally broken but I am able to get 100% recycled Buffers with a custom

Re: Patch for IoBuffer management in IoProcessor 2.0.7

JIRA location? On Thu, Jan 3, 2013 at 11:35 AM, Emmanuel Lécharny elecha...@gmail.comwrote: Le 1/3/13 5:12 PM, Jon V. a écrit : Sorry that this is the entire file. I don't have it checked out with subversion to create typical patches. Basically, there are two places where I call free

Re: Patch for IoBuffer management in IoProcessor 2.0.7

https://issues.apache.org/jira/browse/DIRMINA-929 On Thu, Jan 3, 2013 at 3:24 PM, Arnaud bourree arnaud.bour...@gmail.comwrote: http://issues.apache.org/jira/browse/DIRMINA 2013/1/3 Jon V. sybersn...@gmail.com: JIRA location? On Thu, Jan 3, 2013 at 11:35 AM, Emmanuel Lécharny elecha

Fw: Session Closed

I'm looking for how Mina is determining a close from NIO. Since NIO does not have any direct method to detect closure.(without calling write/read and getting a -1) On Friday, January 18, 2013 at 9:37 AM, Emmanuel Lécharny wrote: Le 1/18/13 2:48 PM, Jon V a écrit : Emmanuel

Re: changes the bytes of transmitting byte Array through apache camel mina UDP

+1 for great analogy. On Friday, March 1, 2013 at 9:12 AM, Emmanuel Lécharny wrote: Le 3/1/13 9:50 AM, Tejas Patel a écrit : Problem : When i transmit the byte array of [56, 34, 0, 4, -79, 1] from one device to another device on using Apache Camel Mina UDP endpoint, at receiver

Re: XML Protocol with initial ASCII handshake

1) You obtain the next filter and call messageReceived there. Mina will automatically wrap the last filter and call the handler. 2) nextFilter.filterWrite(session, new DefaultWriteRequest(out, writeRequest.getFuture(), writeRequest.getDestination())); On Fri, Apr 26, 2013 at 2:45 PM, Max Larsson

Rules

To the MINA mailing list owner. Can we setup a rule requiring the project name in the subject line? e.g. [MINA] , [SSHD]

Re: Does MINA 2.0.4 supports IPV6

Are you using wildcards in the blacklist filter? You could build your own string and pattern based blacklist filter for IPV6 in under 30 minutes. On Fri, Jul 26, 2013 at 3:10 AM, Nitin Phuria nit...@integramicro.comwrote: The issue is registered in JIRA see the below link.

RE: Does MINA 2.0.4 supports IPV6

, Nitin Phuria -Original Message- From: Jon V. [mailto:sybersn...@gmail.com] Sent: Saturday, July 27, 2013 12:27 AM To: users@mina.apache.org; nit...@integramicro.com Subject: Re: Does MINA 2.0.4 supports IPV6 Are you using wildcards in the blacklist filter? You could build your own

Re: Using an Authentication Filter and IoHandlerAdapter

To my knowledge the executor filter does not guarantee any kind of order. This means that you should implement the authentication phase before the executor. Io-prorocol-authentication-executor-handler You cannot lock on out of order messaging without a queue and attempt to re-order the messages.

Re: Using an Authentication Filter and IoHandlerAdapter

. Hunter On 8/1/13 9:54 PM, Ashish wrote: On Fri, Aug 2, 2013 at 7:15 AM, Jon V. sybersn...@gmail.com wrote: To my knowledge the executor filter does not guarantee any kind of order. This means that you should implement the authentication phase before the executor. This is correct

Re: Using an Authentication Filter and IoHandlerAdapter

The major problem with any kind of ordering is the blocking and major performance problems relative to the load. On Aug 1, 2013 9:55 PM, Ashish paliwalash...@gmail.com wrote: On Fri, Aug 2, 2013 at 7:15 AM, Jon V. sybersn...@gmail.com wrote: To my knowledge the executor filter does

Re: User authentication

That is basically how it is done. Io-packet processor-state machine-handler The state machine should only handle one packet at a time. In the message received just do... If ( authenticated ) Next filter.message received( msg ) Else Etc On Aug 7, 2013 3:22 PM, Hunter McMillen

Re: User authentication

for the handler will basically be in messageReceived? And the state machine manages sessionCreated()/**sessionOpened()? Hunter On 8/7/2013 3:29 PM, Jon V. wrote: That is basically how it is done. Io-packet processor-state machine-handler The state machine should only handle one packet at a time

Re: AW: AW: NioSocketConnector/NioSocketAcceptor

Please don't take this the wrong way. The apache Mina site is full of examples and documentation. That should be your primary source of information before asking questions on the email list. Remember, this list covers 5 different projects and your email goes out to a lot of people. I am very

Re: Regarding DIRMINA-912

What is the use case for having different thread pools for read and write events? Writes are already scheduled one they hit the worker. Not like your write process is going to slow down an already threaded read request. On Aug 22, 2013 5:52 PM, Mike McKnight m...@mcknight4.com wrote: I am just

Re: Regarding DIRMINA-912

like this: * codec * inbound-filter (executor filter for all but WRITE events) * write-throttle-filter (overrides filterWrite) * outbound-filter (executor filter for WRITE events only) On 08/23/2013 12:32 PM, Jon V. wrote: What is the use case for having different thread pools for read

Re: Send/Receive IoBuffer

On Mon, Sep 2, 2013 at 5:37 PM, Luciano Coelho luci...@lvcsistemas.comwrote: Hello everybody. I'm learning how to use Mina and my questions are: I solved my problem of having a answer from my server without closing the session like this: System.out.flush(); This really gives me

Re: Send/Receive IoBuffer

and the server answer quickly, the only problem now is this big request that one time goes another time not. Thank's again. LUCIANO -Mensagem original- De: Jon V. [mailto:sybersn...@gmail.com] Enviada em: segunda-feira, 2 de setembro de 2013 19:26 Para: users@mina.apache.org Assunto: Re

Re: Send/Receive IoBuffer

have enough data. On Mon, Sep 2, 2013 at 10:42 PM, Jon V. sybersn...@gmail.com wrote: On Mon, Sep 2, 2013 at 10:21 PM, Luciano Coelho luci...@lvcsistemas.comwrote: Yeah Jon, you're right about flush, the wrong thing is really the IoBuffer. I put the: session.getConfig().setReadBufferSize

Re: Send/Receive IoBuffer

Fix if( buffering == null ) { buffering = ByteBuffer.allocate( some size ); session.setAtribute( SOME_PROPERTY_BUFFER_NAME, buffering ); } On Mon, Sep 2, 2013 at 10:54 PM, Jon V. sybersn...@gmail.com wrote: This is better: I am not going to use the correct API but you can get the idea

Re: Send/Receive IoBuffer

. -Mensagem original- De: Jon V. [mailto:sybersn...@gmail.com] Enviada em: segunda-feira, 2 de setembro de 2013 23:54 Para: users@mina.apache.org Assunto: Re: Send/Receive IoBuffer This is better: I am not going to use the correct API but you can get the idea. Create a Filter called

Re: MINA support for WebSockets?

The http implementation hasn't received any love lately as far as I know. It is not for needing a bit of work. Why not take the code and integrate it yourself? WebSockets are not terribly complicated. On Sep 4, 2013 11:36 AM, Mark Phillips gwailoh2...@yahoo.com wrote: The Users and Dev list

Re: Trouble with messageReceived()

On the client you are writing a hashmap to the session. I don't see a filter which turns the hashmap into a IoBuffer? Only IoBuffer can be written to the socket directly. On Sep 16, 2013 1:12 PM, Karrys, Michael (IS) mike.kar...@ngc.com wrote: I was wondering if someone could look at this and

RE: EXT :Re: Trouble with messageReceived()

LoggingFilter()); But it drops messages. The SumUp example does not seem to drop messages when I changed it to use the ObjectSerializationCodecFactory but I will have to verify that. -Original Message- From: Jon V. [mailto:sybersn...@gmail.com] Sent: Monday, September 16, 2013 12:25

Re: Multithreaded UDP server vs. OrderedThreadPoolExecutor

The first thing I would do would be to disable the executor and try again. You could be encountering some sort of deadlock. On Sep 24, 2013 12:47 PM, Christian Hammers chamm...@netcologne.de wrote: Hello I've written a server for a proprietary UDP protocol that needs to answer with exactly

Re: Multithreaded UDP server vs. OrderedThreadPoolExecutor

Depends on what aspect of asynchronous you are looking for. The sockets are always async. If you have a bunch of ioprocessors then you have concurrency where more than one message can be processed at any given time. On Sep 25, 2013 6:15 AM, Christian Hammers chamm...@netcologne.de wrote: Hello

Re: Unable to open more than 1000 ports in Mina Client

NioSocketAcceptor( number ) creates N threads. 10 is a problem in most cases. The default backlog for a SocketAcceptor in Java is 50. I have mine set to 200 to prevent dropped accept requests. Each time a socket attempts to connect it gets put in the backlog waiting for an accept to be

Re: Unable to open more than 1000 ports in Mina Client

on Linux; there is a significant performance problem with greater than 3 on Windows. On Mon, Sep 30, 2013 at 11:54 AM, Emmanuel Lécharny elecha...@gmail.comwrote: Le 9/30/13 8:52 AM, Jon V. a écrit : NioSocketAcceptor( number ) creates N threads. 10 is a problem in most cases. The default

Re: Unable to open more than 1000 ports in Mina Client

at a time; I am sure it would work normally. On Mon, Sep 30, 2013 at 11:59 AM, Jon V. sybersn...@gmail.com wrote: I've seen this before and it was tightly bound to the Acceptor backlog for me. My desktop is Win 2008 r2 and I can get 18,000 tcp connections in Java. Selectors for some reason are very

Re: IoSession write method thread safe?

for writing, is brutal. I am thinking on including a writing lock as a session attribute in each session that will be initialized on sessionCreated events. Cheers! Alex On Thu, Nov 21, 2013 at 4:28 PM, Jon V. sybersn...@gmail.com wrote: I would imagine this to be true. The normal use case

Re: IoSession write method thread safe?

I would imagine this to be true. The normal use case is request and response. Additional synchronization makes the code run slower in that scenario. I would suggest synchronizing critical paths. I have to synchronize write paths for the same reason where one input may write to dozens of other

Re: Can MINA create proxy servers?

I don't see why not. I wouldn't recommend it however. Proxy servers need to be more bare metal. Nginx has a proxy module you should look at. On Nov 22, 2013 11:59 AM, Andrew Pennebaker apenneba...@42six.com wrote: That would be cool! -- Cheers, Andrew Pennebaker apenneba...@42six.com

Re: Can MINA create proxy servers?

. Would be nice to spin up proxy servers, to test proxy support in the crawler. On Fri, Nov 22, 2013 at 12:02 PM, Jon V. sybersn...@gmail.com wrote: I don't see why not. I wouldn't recommend it however. Proxy servers need to be more bare metal. Nginx has a proxy module you should look

Re: Can MINA create proxy servers?

I was thinking about the Http lib. I didn't remember a proxy in there. Thanks. On Nov 22, 2013 3:32 PM, Emmanuel Lécharny elecha...@gmail.com wrote: Le 11/22/13 9:20 PM, Jon V. a écrit : I imagine you could build a proxy in Mina. However I don't believe any prebuilt proxies exist. May

Re: handling a slow client

You should never assume that all of your data is available. If you work with this, you can handle any speed client. I test everything against simulated invalid packets and random latency. On Nov 25, 2013 9:38 AM, Andrew Pennebaker apenneba...@42six.com wrote: This is a good edge case to

Re: Separate Threads for sending and receiving

Instead of relying on logging; I would setup a stress test to measure completed commands/s and use that as a foundation for trying other procedures. On Sat, Feb 22, 2014 at 7:03 AM, Max Larsson max.lars...@facilityboss.bizwrote: Hi, In my IoHandler i receive the desired message and

Re: How to write byte[] to an IoSession

You are converting an incoming iobuffer to byte[] you need to write an iobuffer. On Mar 12, 2014 11:34 AM, dinesh707 dinesh...@gmail.com wrote: As shown below I can read the byte[] sent from an IoSession without having a protocol decoder. IoBuffer in = (IoBuffer) message; byte[] inBytes =

Re: [Support] Mina 2.0.7 improve reliability when there are network error

TCP works perfectly. If there is a problem. It is elsewhere. Implementing message delivery confirmation is a part of a higher level protocol. If the network/TCP is connected; it will be delivered. In what environment are you expecting TCP sockets to close all the time? Here is what happening

Re: sshd - async interface

You could simply try the write or read on an interval. There are a few scenarios where aio is use able without NIO. Primarily when you use blocking read and scheduled non blocking writes. On Apr 19, 2014 9:58 AM, Alon Bar-Lev alon.bar...@gmail.com wrote: On Sat, Apr 19, 2014 at 3:52 PM, Jon V

Re: Help with so many concurrent connections

This isn't really a Mina question. Seems like you need to work on some tests to find out what parts are the bottleneck. I don't know why a buffer would mean faster inserts. Buffers are good to offload the tasks to other threads and that has its own performance advantages depending on how your

Re: Performance issue - NioSocketAcceptor

Emmanuel, is there a line test tool for Mina? I ask because seems like the Jedi Jin here has a throughput problem. Jin, Have you tried benchmarking your code directly? In all reality I process 30k sockets and 500mb/s using raw TCP in java so if there is a problem it is either your code or Mina.

Re: Performance issue - NioSocketAcceptor

number of messages. On Aug 15, 2014 2:00 AM, Emmanuel Lécharny elecha...@gmail.com wrote: Le 15/08/14 05:41, Jon V. a écrit : Emmanuel, is there a line test tool for Mina? I ask because seems like the Jedi Jin here has a throughput problem. Something like http://blog.softlayer.com/2011/using

Re: Is IoSession.write() thread-safe in mina trunk branch?

It is my understanding that the write is occurring on the the selector loop. This occurs on the same selector loop every time and therefore is thread-safe. When you call write the params are encapsulated and triggered for execution on the select loop. It does not occur in the same stack as the

Re: Mina server not responsive

We are going to need a lot more. I would first use visual VM and figure if any threads are running 100% and grab stack traces. On Mar 21, 2015 3:03 PM, Jianbao (Jim) Tao jianbao@gmail.com wrote: Hi, I am facing a issue as follows. We have a proxy server developed with Mina. Recently

Re: Session Sharing between applications

I am going to take a little leap and say that Java is probably not the best to develop a proxy application. Mina's design is also more optimized for request response type situations and may introduce some unnecessary latency into the proxy. Otherwise, it is just as Emmanuel said. You write all

Re: Virtual IP Load Balancer

This is not the correct place to discuss this. Maybe try stack overflow. On Wed, Mar 25, 2015 at 4:59 AM, rkreddy.bogati rkreddy.bog...@gmail.com wrote: @Ukslim, Thanks for your response. Let me explain the details requirement. I have one IP Address (that is not connected to any physical

Re: Virtual IP Load Balancer

the job. On Wed, Mar 25, 2015 at 1:06 PM Jon V. sybersn...@gmail.com wrote: This is not the correct place to discuss this. Maybe try stack overflow. On Wed, Mar 25, 2015 at 4:59 AM, rkreddy.bogati rkreddy.bog...@gmail.com wrote: @Ukslim, Thanks for your response. Let me explain

Re: Mina server not responsive

Glad you figured it out. On Mar 22, 2015 6:39 PM, Jianbao (Jim) Tao jianbao@gmail.com wrote: Thank you, Jon and Emmanuel. I tried adding a logging filter and all. I finally found the root cause by taking a thread dump. It was due to a deadlock in our business logic code. I tried

Re: question

This isn't necessarily something that would be the responsibility of Mina. Just as you developed your own game server, you will have to add any kind of clustering features yourself. On Mar 24, 2015 11:16 AM, antares anta...@virtuasport.it wrote: Hi Folks, i created a game server and i would

Re: How to debug ghost sessions

Ghost sessions are a result of a broken TCP connection. It is possible that the client lost internet connection and was unable to correctly terminate the TCP session. Therefore, your application will never get a TCP close until the default idleness timer invokes or you use Mina's internal

Re: EOFException when file gets modified during sendFile

Maybe the file should be locked until send is complete. Really shouldn't allow sending of a file while it is being modified. On May 15, 2015 5:31 PM, Charlie Kim charl...@yahoo-inc.com.invalid wrote: Hi there, I'm seeing EOFException when file gets modified during sendFile command. Here is

Re: How to Detect Wired Network Disconnect in Apache Mina

Write and read is only the real way to detect disconnect. Sockets can be valid and live on after a disconnect and reconnect. When a network cable is removed, you may get a disconnect after the OS internal timeout. Otherwise you have to rely on your own idleness timer. On Apr 17, 2015 6:28 AM,

Re: How to Detect Wired Network Disconnect in Apache Mina

can take decision. On Fri, Apr 17, 2015 at 6:00 PM, Jon V. sybersn...@gmail.com wrote: Write and read is only the real way to detect disconnect. Sockets can be valid and live on after a disconnect and reconnect. When a network cable is removed, you may get a disconnect after the OS

Re: ProtocolCodecFilter threading model

You also cannot process a linear buffer simultaneously. TCP is linear and reading the messages must be linear. After you have a list of messages then you may execute them in parallel but the line decoding of the buffer must be single threaded. The decoder operates on an inbound buffer and

Re: ProtocolCodecFilter threading model

reams > in parallel? > 7 окт. 2015 г. 10:18 PM пользователь "Jon V." <sybersn...@gmail.com> > написал: > > > You also cannot process a linear buffer simultaneously. TCP is linear > and > > reading the messages must be linear. After you have a list of m

Re: Memory leak

Emmanuel is correct. You have to manage that backlog. Mina provides write request handlers so you get notifications when the data is written. If you detect that data isn't writing you should take steps to mitigate sending data to that session. On Sep 1, 2015 12:01 PM, "Emmanuel Lécharny"

Re: where is ssl filter source code?

Here https://git1-us-west.apache.org/repos/asf?p=mina.git;a=summary You can choose which HEAD you want. trunk is 2 years old so I wouldn’t look at that. Here is the “2.0” HEAD https://git1-us-west.apache.org/repos/asf?p=mina.git;a=tree;h=refs/heads/2.0;hb=refs/heads/2.0 Its 4 days old. On

Re: where is ssl filter source code?

Here is the browser for the filters package https://git1-us-west.apache.org/repos/asf?p=mina.git;a=tree;f=mina-core/src/main/java/org/apache/mina/filter;hb=d0263e8a00cf6e60d18317d1e8899554c53083bc On Tue, Dec 22, 2015 at 4:46 AM, Emmanuel Lécharny wrote: > Le 22/12/15

Mina workflow

Its been a while since I’ve used Mina and I forgot how the processor model works. Can anyone confirm the diagram is correct?

Re: Deadlock when using SslFilter and ProxyFilter together

I would have thought the reason SSL Filter needs to be before Proxy is because the SSL handshaking from the client side would get proxied away from the SSL Filter. On Wed, Feb 10, 2016 at 12:57 PM, Emmanuel Lécharny wrote: > Le 10/02/16 18:16, Norbert Irmer a écrit : > >>

Re: Deadlock when using SslFilter and ProxyFilter together

Its my understanding that Read and Write operations are thread-safe and should never cause deadlocks. (not sure why there are locks in those filters) Also, I don’t think SSL should be after Proxy. You cannot initiate an SSL session if you proxy the setup routine. What is the goal in this setup?

Re: SslFilter with autoStart=false

to me more like a fundamental flow problem of filter modifications than a design issue with SSL filter. On Thu, Jan 28, 2016 at 10:45 AM, Emmanuel Lécharny <elecha...@gmail.com> wrote: > Le 28/01/16 15:37, Jon V. a écrit : > > Sorry, I hit enter and my email sent. > >

Re: Can we get message attempted to write

After the message is converted to ByteBuffer is the WriteRequest triggered or it a child WriteRequest created for the ByteBuffer? That would define if this kind of callback would be possible. On Tue, Feb 2, 2016 at 7:35 AM, Emmanuel Lécharny wrote: > Le 02/02/16 12:10,

Re: SslFilter with autoStart=false

Sounds to me like someone has a secondary filterchain they are trying to prime artificially. So then maybe this isn’t necessary? On Thu, Jan 28, 2016 at 9:34 AM, Emmanuel Lécharny <elecha...@gmail.com> wrote: > Le 28/01/16 15:30, Jon V. a écrit : > > In what scenario would

Re: SslFilter with autoStart=false

orkarounds. On Thu, Jan 28, 2016 at 9:37 AM, Jon V. <sybersn...@gmail.com> wrote: > Sounds to me like someone has a secondary filterchain they are trying to > prime artificially. > > So then maybe this isn’t necessary? > > > > On Thu, Jan 28, 2016 at 9:34 AM, Emmanue

Re: SslFilter with autoStart=false

In what scenario would the SSL filter be attached but the session has no filterchain On Thu, Jan 28, 2016 at 7:40 AM, Emmanuel Lécharny wrote: > Le 28/01/16 13:33, Norbert Irmer a écrit : > > Your idea, adding a method "public void initiateHandshake(IoSession > session)"

Re: Why are high ports used by SFTP server implementation?

I’ll try to simplify; TCP (the protocol) can have up to 0x ports or 65535 Ports only need to be reserved for server services and outbound connections. They both use the same pool size of 65535. Each OS type allocates different range for user-space applications. For linux its around 32768

Re: one question with max connections

Probably creating the connections too quickly? make the TCP backlog larger. Default is 10 I believe; I use 100 in production. On Wed, Feb 17, 2016 at 6:44 AM, Liu, Li (Nokia - CN/Hangzhou) < li@nokia.com> wrote: > Hi Mina > > I try to use Mina to handle multiple TCP’s connections > > But

Re: re : Server Crash

I retract my previous statement after pulling the source for CircularQueue. Guess I'm used to doing illegal things in Java I forget that most code isn't like that. On Thu, Apr 27, 2017 at 8:38 AM, Jon V. <sybersn...@gmail.com> wrote: > It looks like a Mina crash to me. Probably