Re: re : Server Crash

[Jon V.]

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 some kind of Unsafe access.
>
> Can you tell us the exact version of Mina?
>
> On Thu, Apr 27, 2017 at 8:20 AM, Emmanuel Lécharny <elecha...@gmail.com>
> wrote:
>
>> Hi !
>>
>>
>> seems like I have hard time moderating your message to the users
>> mailinglist... Never mind. here is a copy :
>>
>>
>> > We have a standalone application. It crashed suddenly. The log contains
>> a
>> > class which is found in mina-core jar file. I have attached the crash
>> > report. Please let me know if you could find out any root cause for the
>> > crash. Thank you. server_crash_report.txt
>> >
>> <http://apache-mina.10907.n7.nabble.com/file/n52619/server_c
>> rash_report.txt>
>>
>> First of all, this is a JVM crash, not a MINA crash The fact that the
>> crash occurred while the exacution was in a mina call is incidental.
>>
>> I would strongly to switch to a more recent version of Java, the one you
>> use is quite antiquated and not supported anymore (4 years old, the last
>> supported Java 6 version from Oracle, which has issued 16 released for Java
>> 6).
>>
>> Java 8 is the current version, Java 7 has been EOLed 2 years ago.
>>
>>
>> --
>> Emmanuel Lecharny
>>
>> Symas.com
>> directory.apache.org
>>
>>
>

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

[Jon V.]

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 ports.  When creating outbound connections the OS will
randomly select an unused port.  This is this high port number.

A connected TCP socket looks like this: local:45223 (ephemeral) <->
remote:22 (fixed)

What you are seeing is the port number of the client.  That number is
allocated on their machine and not the server.

Look up TCP on Wikipedia.

On Tue, Feb 23, 2016 at 8:52 PM, David Hoffer  wrote:

> Hum, that's not entirely clear to me.  The first link says...
>
> 'A TCP/IPv4 connection consists of two endpoints, and each endpoint
> consists of an IP address and a port number.  Therefore, when a client user
> connects to a server computer, an established connection can be thought of
> as the 4-tuple of (server IP, server port, client IP, client port).
> Usually three of the four are readily known -- client machine uses its own
> IP address and when connecting to a remote service, the server machine's IP
> address and service port number are required.
>
> What is not immediately evident is that when a connection is established
> that the client side of the connection uses a port number.  Unless a client
> program explicitly requests a specific port number, the port number used is
> an *ephemeral* port number.  Ephemeral ports are temporary ports assigned
> by a machine's IP stack, and are assigned from a designated range of ports
> for this purpose.
> In our case the server is configured to listen on port 22 and the client
> connects to port 22 so isn't that fixing the port on both sides at port
> 22?  Are you saying that although port 22 is the logical port used on both
> systems, that in reality a different port is used on the client to connect
> to the server?  We are using SSH only here I understand that only used port
> 22.
>
> Regarding the second link is that for FTP or also for SFTP?  I know FTP
> uses passive ports and so does FTPS but we are only using SFTP, e.g. file
> transfer as part of SSH.
>
> Do those links really describe my situation?  Or are those high ports
> created on the server so it can hand off work so it can listen on 22
> again?  E.g. is it using separate ports to communicate with clients instead
> of multiple threads on same port?
>
> Its not clear to me yet, trying to understand.
>
> -Dave
>
>
>
>
> On Tue, Feb 23, 2016 at 4:32 PM, Chad Beaulac  wrote:
>
> > Hey Dave,
> >
> > Listener servers hand off to ephemeral ports.
> > http://www.ncftp.com/ncftpd/doc/misc/ephemeral_ports.html
> > You need ephemeral ports so a server can start listening on port 22 again
> > while something else is happening.
> >
> > Look here for some configuration options.
> >
> https://mina.apache.org/ftpserver-project/configuration_passive_ports.html
> >
> > -Chad
> >
> >
> > On Tue, Feb 23, 2016 at 3:09 PM, David Hoffer 
> wrote:
> >
> > > We are using SSHD in an application to create an embedded SFTP server
> > which
> > > works fine.  Our clients connect on port 22 and we don't have any issue
> > > with that.
> > >
> > > The problem/question is that our IA folks are complaining that our app
> > also
> > > listens on what appear to be random high ports.  E.g. I see this in our
> > > logs.
> > >
> > > Session username@/127.0.0.1:58118 authenticated
> > > Server session created from /127.0.0.1:58132
> > > Server session created from /127.0.0.1:58139
> > > Server session created from /127.0.0.1:58157
> > >
> > > I see these later log statements are coming from IoSession in
> > > ServerSessionImpl but I don't call this in my code so must be part of
> the
> > > SSHD/MINA framework.
> > >
> > > Why are these high ports being used and do we need them?  If not needed
> > for
> > > SFTP server how can I disable?  If they are needed, why and can I
> control
> > > the exact ports that are used?
> > >
> > > -Dave
> > >
> >
>

Re: one question with max connections

[Jon V.]

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 there are some connections that cannot be handled by Mina under lots
> of connections was opened as below :
> Eg, I try to open 400 TCP’s connections, there are always losing 4 or 5
> connections
>
>
>
> [root@twslave2 test]# netstat -natp | grep 62397
> tcp   *22*  0 :::10.70.81.114:65534   :::
> 10.70.81.106:62397   ESTABLISHED -
> [root@twslave2 test]# netstat -natp | grep 62398
> tcp   22  0 :::10.70.81.114:65534   :::10.70.81.106:62398
> ESTABLISHED -
> [root@twslave2 test]# netstat -natp | grep 62399
> tcp   22  0 :::10.70.81.114:65534   :::10.70.81.106:62399
> ESTABLISHED -
> [root@twslave2 test]# netstat -natp | grep 62400
> tcp   22  0 :::10.70.81.114:65534   :::10.70.81.106:62400
> ESTABLISHED -
> [root@twslave2 test]# netstat -natp | grep 62401
> tcp   22  0 :::10.70.81.114:65534   :::10.70.81.106:62401
> ESTABLISHED -
> [root@twslave2 test]# netstat -natp | grep 62396
> tcp0  0 :::10.70.81.114:65534   :::10.70.81.106:62396
> * ESTABLISHED 12152/java*
> [root@twslave2 test]#
>
> Can you tell me why?
>
>

Re: Deadlock when using SslFilter and ProxyFilter together

[Jon V.]

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 :
> >> Ok, but you can insert teh SslFilter *before* the proxyFilter. Once you
> >> get the ProxyConnector instance, get the connector and do :
> >
> > Well, I am using a patched QuickFixJ version, in which the control flow
> is a bit different.
> >
> > Here is an excerpt from the code they are using:
> >
> >
> >   CompositeIoFilterChainBuilder ioFilterChainBuilder = new
> CompositeIoFilterChainBuilder(
> > userIoFilterChainBuilder);
> >
> > boolean useProxy = proxyType != null && proxyPort > 0 &&
> > socketAddresses[0] instanceof
> InetSocketAddress;
> >
> >
> > SslFilter sslFilter = null;
> > if (sslEnabled) {
> > boolean autoStart = !useProxy;
> > sslFilter = installSslFilter(ioFilterChainBuilder,
> autoStart);
> > }
> >
> >
>  ioFilterChainBuilder.addLast(FIXProtocolCodecFactory.FILTER_NAME,
> > new ProtocolCodecFilter(new
> FIXProtocolCodecFactory()));
> >
> > if (useProxy) {
> > SocketConnector socketConnector = new
> NioSocketConnector();
> >
>  socketConnector.setFilterChainBuilder(ioFilterChainBuilder);
> > ioConnector = ProtocolFactory.createIoProxyConnector(
> socketConnector,
> > (InetSocketAddress) socketAddresses[0], new
> InetSocketAddress(proxyHost,
> > proxyPort), proxyType, proxyVersion,
> proxyUser, proxyPassword,
> > proxyDomain, proxyWorkstation);
> > } else {
> > ioConnector =
> ProtocolFactory.createIoConnector(socketAddresses[0]);
> > ioConnector.setFilterChainBuilder(ioFilterChainBuilder);
> > }
> >
> > ioHandler = new InitiatorIoHandler(fixSession, sslFilter,
> networkingOptions, eventHandlingStrategy);
> > }
> >
> > private SslFilter installSslFilter(CompositeIoFilterChainBuilder
> ioFilterChainBuilder,
> >boolean autoStart)
> > throws GeneralSecurityException {
> > SslFilter sslFilter = new
> SslFilter(SSLContextFactory.getInstance(keyStoreName,
> > keyStorePassword.toCharArray()), autoStart);
> > if (enableProtocole != null)
> > sslFilter.setEnabledProtocols(enableProtocole);
> > if (cipherSuites != null)
> > sslFilter.setEnabledCipherSuites(cipherSuites);
> > sslFilter.setUseClientMode(true);
> > ioFilterChainBuilder.addLast(SSLSupport.FILTER_NAME,
> sslFilter);
> > return sslFilter;
> > }
> >
> >
> > Sure, with a lot of time, I could possibly rewrite the code, to swap the
> positions of the filters in the filterchain.
> >
> > But  I see little sense in it, since the code is working. The only
> problem I have, is that I get this deadlock after a few minutes or so.
> How can you say the code is working when you get a deadlock ?
>
> As Jon said, having the ProcyFilter *before* the SSLFilter in the chain
> is not the right thing. The reason being that you may perfectly get
> fragmented TCP buffers, and the proxy might not be able to deal with
> de-fragmentation when it comes to encrypted data.
>
> I suggest that you patch teh code to first call the installSslFilter
> method after the proxy has been created, and in the
> installSslFiltermethod, use addFirst() instead of addLast().
>
> Otherwise, you can also check with the latest trunk revision I built
> today to see if it's any better (on http://people.apache.org/~elecharny)
>
>

Re: Deadlock when using SslFilter and ProxyFilter together

[Jon V.]

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?

Please send us the deadlock thread trace.(You can get it from VisualVM)

On Tue, Feb 9, 2016 at 11:59 AM, Norbert Irmer  wrote:

> Hello,
>
> I detected a deadlock when using Proxies and Ssl together.
>
> The SslFilter and the ProxyFilzter both have synchronized blocks, and when
> sending
> and receiving messages you get deadlocks.
>
> The SslFilter hangs in "sslLock.lock()" in
> "SslHandler.flushScheduledEvents()",
> while the ProxyFilter hangs in "synchronized(handler)" in
> "ProxyFilter.writeData".
>
> That sending and receiving messages with both filters simultaneously leads
> to a deadlock
> is clear, since both filters have locks and data flow direction is
> reversed when reading and writing:
>
> Read direction:   Socket  --> ProxyFilter   ---> SslFilter  -->
> FixProtocol   --> Application
> Write direction:  Socket <-- ProxyFilter  <---  SslFilter  <--
> FixProtocol  <-- Application
>
>
>
>

Re: Can we get message attempted to write

[Jon V.]

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, Abhijit Bhatode a écrit :
> > Hi,
> >
> > We are using MINA 2.0 M1.
> You should switch to 2.0.11 which has been released last wek... 2.0.0-M1
> is 8 years old !
>
> > Is there a way to get details of message when an
> > exception occurs during write?
>
> It depends...
>
> There are many potential exception you can get, and I suspect you wnat
> to get some information in the exceptionCaught() method in your IoHandler.
>
> If you are lucky, the message is still in the write queue. You can get
> it by using :
>
> WriteRequest writeRequest = session.getWriteRequestQueue().poll(
> session );
>
> if (writeRequest != null) {
> Object message = writeRequest.getMessage();
> }
>
> Otherwise, check the cause.
>
> One other possibility is to store in your session the messages that you
> are sending before writing them, and in the exceptionCaught() method,
> you can most certainly check on the session to see what is the message
> that was the cause. This is, again, not guaranteed, because you may have
> many messages waiting to be sent when you get the exception.
>
>
> Keep in mind that this is asynchonous, so the exceptionCaught event can
> be received way after you tried to write a message.
>
> Last, not least, you will receive a messageSent() event for every single
> message that has successfully sent. You can also use this information to
> detect what is the message causing the problem.
>
>
> In any case, as I said, there are many reson why you would receive an
> exception, and a lot of them might be totally disconnect to the fact
> that you tried to write a message.
>
>
> Wht is your use case ?
>
>
>

Re: SslFilter with autoStart=false

[Jon V.]

I understand the tendency for defensive style design.

Auto start option should probably be removed from the SSL Filter.  If OP
wishes to construct the chain at Open() time then calling SSL Filter.open()
manually should initiate the handshake.

  @Override
public void proxySessionOpened(IoSession session) throws Exception {
log.info("proxySessionOpened called");
if (sslFilter != null) {
IoFilter.NextFilter nextFilter = session.getFilterChain().
getNextFilter(sslFilter);
sslFilter.sessionOpened(next, session)
}
}

This seems 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.
> >
> > IoFilterChain filterChain = session.getFilterChain();
> >
> > if (filterChain == null) {
> > throw new SSLException("No filter chain");
> > }
> >
> >
> > This isn’t necessary.  I think we need to know more about OPs goals
> before
> > adding hacky workarounds.
>
> Hmm, yeah, it's probably over-protective.
>
> With respect to such check, I should first have added a if (session ==
> null), and not having such a check makes teh second check a bit spurious.
>
> We can drop it and only have :
>
> public void initiateHandshake(IoSession session) throws SSLException {
> IoFilter.NextFilter nextFilter =
> session.getFilterChain().getNextFilter(SslFilter.class);
>
> if (nextFilter == null) {
> throw new SSLException("No SSL next filter in the chain");
> }
>
> initiateHandshake(nextFilter, session);
> }
>
> For teh same reason, I don't see any use case where we can't have a
> nextFilter, so the if (nextFilter == null) is also spurious.
>
>
> I may have been hit too many times by pervasive NPE, and I try to catch
> them before they happen...
>

Re: SslFilter with autoStart=false

[Jon V.]

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 the SSL filter be attached but the session has no
> > filterchain
>
> None that I know.
>
> It would not make any sense.
>

Re: SslFilter with autoStart=false

[Jon V.]

Sorry, I hit enter and my email sent.

IoFilterChain filterChain = session.getFilterChain();

if (filterChain == null) {
throw new SSLException("No filter chain");
}


This isn’t necessary.  I think we need to know more about OPs goals before
adding hacky workarounds.

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, Emmanuel Lécharny <elecha...@gmail.com>
> wrote:
>
>> Le 28/01/16 15:30, Jon V. a écrit :
>> > In what scenario would the SSL filter be attached but the session has no
>> > filterchain
>>
>> None that I know.
>>
>> It would not make any sense.
>>
>
>

Re: SslFilter with autoStart=false

[Jon V.]

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)" to the SslFilter class,
> > would be perfectly fine for me.
> >
> > I just tested it, and it works fine. Only thing, perhaps, is, that I
> don't know, if session.getFilterChain
> > can return null, so perhaps some null pointer handling is necessary.
> >
> >
> Yes, actually, I have a better implementation that tests corner cases :
>
> public void initiateHandshake(IoSession session) throws SSLException {
> IoFilterChain filterChain = session.getFilterChain();
>
> if (filterChain == null) {
> throw new SSLException("No filter chain");
> }
>
> IoFilter.NextFilter nextFilter =
> filterChain.getNextFilter(SslFilter.class);
>
> if (nextFilter == null) {
> throw new SSLException("No SSL next filter in the chain");
> }
>
> initiateHandshake(nextFilter, session);
> }
>
>
> This is what I'm going to commit.
>

Mina workflow

[Jon V.]

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: where is ssl filter source code?

[Jon V.]

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 Tue, Dec 22, 2015 at 2:06 PM, Vic Cekvenich 
wrote:

> Hi Emanuel,
> Yes that works and matches source.
> The magic is switching to that proper branch.
>
> I like to look under the hood.
> Cheers,
> Vic
>
>
> On Tue, Dec 22, 2015 at 1:46 AM, Emmanuel Lécharny 
> wrote:
>
> > Le 22/12/15 05:30, Vic Cekvenich a écrit :
> > > Hi Ben, I think I should check first, before I say things are missing.
> > > And it does appear that way. For example there are packages such as
> > > sslfilter under filter that show up in docs, yet when you check out the
> > > source code and search for those files, they are not there.
> > >
> > > It is my understanding that this mail list is for the mina project, so
> I
> > > guess I'm contacting the mina project that: there is missing source
> code
> > > that is documented on www in snippets and mina is an open source
> project.
> > > I would very much like to access the full source code documented - as
> > > linked by Emanuel:
> > > http://mina.apache.org/mina-project/xref/org/apache/mina/filter
> > > Cheers,
> > > Vic
> >
> > We use git at MINA. Youc an get the full sources doing :
> >
> > git clone http://git-wip-us.apache.org/repos/asf/mina.git mina
> >
> > Now, what you get is the *full* MINA repo, and that includes MINA 3.0,
> > probably not what you want. Here is what you get when you run this
> command
> > :
> >
> > $ git tag
> > 0.8.3
> > 0.8.4
> > ...
> > 2.0.5
> > 2.0.6
> > 2.0.7
> > 2.0.8
> > 2.0.9
> > 3.0.0-M1
> > 3.0.0-M2
> > mina-parent-3.0.0-M1
> >
> >
> > You need to checkout the MINA 2.0.10 tag :
> >
> > $ git checkout -b 2.0.10 2.0.10
> > Switched to a new branch '2.0.10'
> >
> >
> > That will allow you to see the full source used by this tag, and to
> > retreive the file you are looking at :
> >
> > $ find . -name SslFilter.java
> > ./mina-core/src/main/java/org/apache/mina/filter/ssl/SslFilter.java
> >
> > I'll update the http://mina.apache.org/mina-project/sources.html page to
> > reflect that.
> >
> > Thanks !
> >
> >
>

Re: where is ssl filter source code?

[Jon V.]

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 05:30, Vic Cekvenich a écrit :
> > Hi Ben, I think I should check first, before I say things are missing.
> > And it does appear that way. For example there are packages such as
> > sslfilter under filter that show up in docs, yet when you check out the
> > source code and search for those files, they are not there.
> >
> > It is my understanding that this mail list is for the mina project, so I
> > guess I'm contacting the mina project that: there is missing source code
> > that is documented on www in snippets and mina is an open source project.
> > I would very much like to access the full source code documented - as
> > linked by Emanuel:
> > http://mina.apache.org/mina-project/xref/org/apache/mina/filter
> > Cheers,
> > Vic
>
> We use git at MINA. Youc an get the full sources doing :
>
> git clone http://git-wip-us.apache.org/repos/asf/mina.git mina
>
> Now, what you get is the *full* MINA repo, and that includes MINA 3.0,
> probably not what you want. Here is what you get when you run this command
> :
>
> $ git tag
> 0.8.3
> 0.8.4
> ...
> 2.0.5
> 2.0.6
> 2.0.7
> 2.0.8
> 2.0.9
> 3.0.0-M1
> 3.0.0-M2
> mina-parent-3.0.0-M1
>
>
> You need to checkout the MINA 2.0.10 tag :
>
> $ git checkout -b 2.0.10 2.0.10
> Switched to a new branch '2.0.10'
>
>
> That will allow you to see the full source used by this tag, and to
> retreive the file you are looking at :
>
> $ find . -name SslFilter.java
> ./mina-core/src/main/java/org/apache/mina/filter/ssl/SslFilter.java
>
> I'll update the http://mina.apache.org/mina-project/sources.html page to
> reflect that.
>
> Thanks !
>
>

Re: ProtocolCodecFilter threading model

[Jon V.]

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
multi-threading this process will cause the buffers to complete out of
order.  There is also no guarantee that the inbound buffer contains a
complete message.  In this scenario the tail end of a message could be
processed before the previous part which makes the whole thing totally
broken without locks and locks would just eat more cpu with zero gain.
On Oct 7, 2015 2:53 PM, "Emmanuel Lécharny"  wrote:

> Le 07/10/15 20:44, Марат Гайнуллин a écrit :
> > Dear sirs!
> >  I am happy to know, that DIRMINA-1013 has been fixed.
> > I have a question. I think, that while decoding TCP messages in separate
> > sessions, server SHOULD do it in parallel. Why your last comment on
> > DIRMINA-934 was about not to put executor filter before decoder in TCP?
> Because doing so will put your server atrisk. Note that you already have
> more than one thread processing incoming messages (IoProcessors) and TCP
> does not guarantee you that a message will be delivered in one piece.
> Now, let's assume you have an exectir *before* the decoder : then for
> one specific session, more than one thread might be used to decode a
> message, which means that potentially, for fragmented messages, more
> than one thread will be in charge of decoding fration of a message. this
> simply does not fly...
>
> For separate sessions, there is absolutely no problem, because each
> session might be handled by a different IoProcessor.
>
> Performance wise, you might object that while a message is being
> decoded, then another session could be blocked because the thread is
> busy decoding (as the two sessions are handled by the same IoProcessor
> thread). that would be true, except that when an IoProcessor thread is
> busy decoding, it uses 100% of the CPU of the core it run on, so there
> is no 'room' for some other exacution. So having an executor before the
> decoder is simply useless, unless you haven't declared as many
> IoProcessors as you have core on your server.
>
> The only case where it's annoying is when your decoder is accessing
> external resources and has to wait for them : unlikely...
> Hope it helps !
>
>

Re: ProtocolCodecFilter threading model

[Jon V.]

It would use more memory and since there is already 2x IoProcessors for cpu
cores there would be no speedup. In fact it would lead to more memory
fragmentation/cache misses and require one thread for every session.

Basically, it would be a bad idea.
On Oct 7, 2015 3:36 PM, "Марат Гайнуллин" <mgaynul...@altsoft.biz> wrote:

> Dear sirs!
> It is completely clear for me, that decoding of a buffer must be linear
> and tcp messages fragments must be decoded in a serial manner. My question
> is about different thing. Why not to decode SEPARATE sessions' data streams
> 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 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
> > multi-threading this process will cause the buffers to complete out of
> > order.  There is also no guarantee that the inbound buffer contains a
> > complete message.  In this scenario the tail end of a message could be
> > processed before the previous part which makes the whole thing totally
> > broken without locks and locks would just eat more cpu with zero gain.
> > On Oct 7, 2015 2:53 PM, "Emmanuel Lécharny" <elecha...@gmail.com> wrote:
> >
> > > Le 07/10/15 20:44, Марат Гайнуллин a écrit :
> > > > Dear sirs!
> > > >  I am happy to know, that DIRMINA-1013 has been fixed.
> > > > I have a question. I think, that while decoding TCP messages in
> > separate
> > > > sessions, server SHOULD do it in parallel. Why your last comment on
> > > > DIRMINA-934 was about not to put executor filter before decoder in
> TCP?
> > > Because doing so will put your server atrisk. Note that you already
> have
> > > more than one thread processing incoming messages (IoProcessors) and
> TCP
> > > does not guarantee you that a message will be delivered in one piece.
> > > Now, let's assume you have an exectir *before* the decoder : then for
> > > one specific session, more than one thread might be used to decode a
> > > message, which means that potentially, for fragmented messages, more
> > > than one thread will be in charge of decoding fration of a message.
> this
> > > simply does not fly...
> > >
> > > For separate sessions, there is absolutely no problem, because each
> > > session might be handled by a different IoProcessor.
> > >
> > > Performance wise, you might object that while a message is being
> > > decoded, then another session could be blocked because the thread is
> > > busy decoding (as the two sessions are handled by the same IoProcessor
> > > thread). that would be true, except that when an IoProcessor thread is
> > > busy decoding, it uses 100% of the CPU of the core it run on, so there
> > > is no 'room' for some other exacution. So having an executor before the
> > > decoder is simply useless, unless you haven't declared as many
> > > IoProcessors as you have core on your server.
> > >
> > > The only case where it's annoying is when your decoder is accessing
> > > external resources and has to wait for them : unlikely...
> > > Hope it helps !
> > >
> > >
> >
>

Re: Memory leak

[Jon V.]

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"  wrote:

> Le 01/09/15 17:26, Leonardo D'Alimonte a écrit :
> > Hello Emmanuel,
> > actually it seems as if the Queue data structure continues to
> > increase..message after message.
> > We cannot check if the our device receive the message, the implemented
> > protocol doesn't allow this, do you mean checking if the session goes in
> an
> > IDLE state? I can add override of method "sessionIdle(IoSession,
> > IdleStatus) to our HandlerAdapter..
>
> If the remote device brutally disconnected from the server, you will
> *not* be able to know about it. That means you will keep sending data to
> this inexistant client, and as this client is not there to read those
> messages, they will just accumulate in memory.
>
> You have to manage the idle state of your sessions : after a delay, you
> might want to kill the session, which will remove all the pending messages.
>
> Another option : when you send a message to a client, you know it has
> been actually sent when you receive the messageSent event. If you don't
> get this event, that means your client is not reading the messages. A
> smart way to mitigate what we call 'slow readers' is to check that each
> message written into a session has been sent and if not, kill the
> session (or kill it after a number of written messages for which you
> havent received a messageSent event).
>
>

Re: EOFException when file gets modified during sendFile

[Jon V.]

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 the scenario.  Client receives the size of the file as 1000.
 Then, before inputstream is read, file gets modified which the file size
 becomes greater that what it was specified and this causes read to fail.

 debug2: exec request accepted on channel 0
 Sink: C0644 1000 foo.txt

 debug2: channel 0: read=0 rfd 5 len 0
 debug2: channel 0: read failed
 debug2: channel 0: close_read
 debug2: channel 0: input open - drain
 debug2: channel 0: ibuf empty

 here is the code that writes the length.

 https://github.com/apache/mina-sshd/blob/master/sshd-core/src/main/java/org/apache/sshd/common/scp/ScpHelper.java#L505
 Could we use available() method of an InputStream and use that instead
 when writing C0644 1000 foo.txt?I've tried this workaround and seems to
 work for me.
 Thanks

Re: How to debug ghost sessions

[Jon V.]

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 idleness to close sessions.

On Mon, May 11, 2015 at 2:33 AM, 白玉雄 ba...@17car.com.cn wrote:

 Hi,
 I print all attributes of a bug session as follow,  Mina processes the
 last request at 2015-05-11 01:58:34, then stucked and no request is
 processed afterwards.


 namevalue
 getWrittenBytesThroughput0.0
 getLastIoTime2015-05-11 02:23:11
 getReadBytes5971
 getReadBytesThroughput0.0
 isConnectedtrue
 getLastWriterIdleTime2015-05-11 01:58:34
 isWriterIdlefalse
 getReaderIdleCount0
 getLocalAddress/xx.xx.xx.xx:
 getWriterIdleCount0
 getWrittenMessagesThroughput0.0
 ID3233
 getServiceAddress/0.0.0.0:x
 getReadMessages30
 getLastReadTime2015-05-11 02:08:48
 isSecuredfalse
 isReaderIdlefalse
 getLastReaderIdleTime2015-05-11 01:58:34
 getWrittenBytes2221
 getRemoteAddress/xx.xx.xx.xx:x
 getReadMessagesThroughput0.0
 isBothIdlefalse
 isReadSuspendedfalse
 getScheduledWriteMessages-39
 getLastWriteTime2015-05-11 02:23:11
 getCreationTime2015-05-11 01:58:34
 getBothIdleCount0
 getLastBothIdleTime2015-05-11 01:58:34
 isWriteSuspendedfalse
 isClosingfalse
 getWrittenMessages39

 -- Original --
 From:  白玉雄ba...@17car.com.cn;
 Date:  Wed, May 6, 2015 02:47 PM
 To:  usersusers@mina.apache.org;

 Subject:  Re: How to debug ghost sessions


 I am sure that idle event is fired. After no heartbeat for 900 seconds ,
 session was closed by server.

 -- Original --
 From:  Emmanuel Lécharnyelecha...@gmail.com;
 Date:  Wed, May 6, 2015 02:31 PM
 To:  usersusers@mina.apache.org;

 Subject:  Re: How to debug ghost sessions


 Le 06/05/15 05:24, 白玉雄 a écrit :
  Hi all,
  My mina server neither process session request nor disconnect the
 session as idle. and this happens only on few of sessions, most of them
 works well.
 
 
  Some Code:
  public void run(int port) throws IOException {
this.acceptor = new NioSocketAcceptor();
this.acceptor.setReuseAddress(true);
 
 
this.acceptor.getFilterChain().addLast(codec,new
 ProtocolCodecFilter(new ClwMessageProtocolCodecFactory()));
 
 
this.acceptor.getFilterChain().addLast(logger, new
 LoggingFilter());
this.acceptor.getFilterChain().addLast(threadPool, new
 ExecutorFilter(new
 UnorderedThreadPoolExecutor(PropertiesHander.getInt(MAXIMUM_POOL_SIZE,
 16;
 
 
this.acceptor.getSessionConfig().setReadBufferSize(512);
this.acceptor.getSessionConfig().setReceiveBufferSize(512);
// set idle time, default 900 seconds
 
  
 this.acceptor.getSessionConfig().setIdleTime(IdleStatus.BOTH_IDLE,PropertiesHander.getInt(SESSION_IDLE_TIME,900));
this.acceptor.setHandler(ioHandler);
this.acceptor.bind(new InetSocketAddress(port));
LOGGER.info(Socket server Listening on port +
 PropertiesHander.getInt(SERVER_PORT, port));
}
 
 
 
  Some phenomenons:
  1、The ghost session can be found by
 this.acceptor.getManagedSessions().get(sessionId)
  2、If force the client disconnect from server and reconnect again, the
 client now works well and request was processed normally.
 
  It seems that the session was stucked.
 
 
  Question:
  Why this happens
 No idea... You aren't providing enough information to know what's going on.

  and How to reappear and debug this?
 Setting the idle time is good, but at some point you must check the
 event that is fired when teh session has been idle for 900 seconds. Are
 you doing that ?

Re: How to Detect Wired Network Disconnect in Apache Mina

[Jon V.]

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, Nikunj Patel aang...@gmail.com wrote:

 Dear Sir,

 I have develop MINA Client which is connected to remote server. I am trying
 to detect Wired Network Disconnect in that but it is not identify it. Also
 session.write is not generating any exception. Before submit if i check
 with session.isConnected then also it return true.

 Can any one help me how I can detect such network disconnection in Apache
 Mina.

 --
 Regards,
 Nikunj Patel.

Re: How to Detect Wired Network Disconnect in Apache Mina

[Jon V.]

I forget how Mina handles it but SocketChannel write and read both generate
exceptions to alert of disconnect.

Mina may abstract that somehow.
On Apr 17, 2015 8:38 AM, Nikunj Patel aang...@gmail.com wrote:

 Is there any way where session.write() can generate any exception and base
 on that we 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
  internal timeout. Otherwise you have to rely on your own idleness timer.
  On Apr 17, 2015 6:28 AM, Nikunj Patel aang...@gmail.com wrote:
 
   Dear Sir,
  
   I have develop MINA Client which is connected to remote server. I am
  trying
   to detect Wired Network Disconnect in that but it is not identify it.
  Also
   session.write is not generating any exception. Before submit if i check
   with session.isConnected then also it return true.
  
   Can any one help me how I can detect such network disconnection in
 Apache
   Mina.
  
   --
   Regards,
   Nikunj Patel.
  
 



 --
 Regards,
 Nikunj Patel.

Re: Virtual IP Load Balancer

[Jon V.]

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 machine),
 which
 will receive the request.
 As soon as request receives it should delegate that request to the given IP
 Addresses (those are attached to any Physical machine) by Round-robin
 algorithm.

 To achieve this I would like to use Open source tool which suites my
 requirement.

 Note:- This requirement for Web-Application developed on JAVA/J2EE
 platform.



 --
 View this message in context:
 http://apache-mina.10907.n7.nabble.com/Virtual-IP-Load-Balancer-tp46678p46698.html
 Sent from the Apache MINA User Forum mailing list archive at Nabble.com.

Re: Virtual IP Load Balancer

[Jon V.]

Should have said Stack Exchange. Meant it in a generic sense.
On Mar 25, 2015 9:18 AM, John Hartnup john.hart...@gmail.com wrote:

 Neither is Stack Overflow the place to discuss it. http://serverfault.com/
 or maybe http://superuser.com/ might be OK. I think my suggestion of
 haproxy will do 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 the details requirement.
  
   I have one IP Address (that is not connected to any physical machine),
   which
   will receive the request.
   As soon as request receives it should delegate that request to the
 given
  IP
   Addresses (those are attached to any Physical machine) by Round-robin
   algorithm.
  
   To achieve this I would like to use Open source tool which suites my
   requirement.
  
   Note:- This requirement for Web-Application developed on JAVA/J2EE
   platform.
  
  
  
   --
   View this message in context:
   http://apache-mina.10907.n7.nabble.com/Virtual-IP-Load-
  Balancer-tp46678p46698.html
   Sent from the Apache MINA User Forum mailing list archive at
 Nabble.com.
  
 

Re: question

[Jon V.]

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 like to have a suggestion about to
 sharing sessions between other game server instances.
 In other words something about clusterizing
 is that possible ?

 Thanks
 Francesco

Re: Mina server not responsive

[Jon V.]

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 upgrading Mina version. However, I ran into a roadblock where Mina
 seemed somehow not receiving new data after the first chunk. It wasn't a
 priority to upgrade Mina at the time so I stopped investigating further.
 Maybe I will bring this here in the future.

 Best regards,
 Jim

 On Sat, Mar 21, 2015 at 5:10 PM, Emmanuel Lécharny elecha...@gmail.com
 wrote:

  Le 21/03/15 20:01, Jianbao (Jim) Tao a écrit :
   Hi,
  
   I am facing a issue as follows.
  
   We have a proxy server developed with Mina. Recently after some changes
  in
   the business logic that hardly seem relevant, the server starts to
 show a
   random behavior frequently, in which it will simply not respond to any
   request. I used curl to manually verify. It appears that the sever is
  able
   to accept connection, but it doesn't seem to accept any data as I
 cannot
   see any logs on the server side.
  
   The version we are using is mina-core 2.0.0-M6.
  
   Right now I am running out of ideas to debug this issue. So, can anyone
   offer some advice about what I can try to investigate this issue? If
 you
   need more details to understand the issue, please, please let me know.
 I
   will do my best to provide them.
 
  I'll suggest you switch first to the latest MINA version (2.0.9).
 
  Regarding your issue, we can't tell anything with the provided infos.
  What I would suggest is that you add a Logger filter to see if the
  server is actually receiving the messages.
 

Re: Mina server not responsive

[Jon V.]

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 after some changes in
 the business logic that hardly seem relevant, the server starts to show a
 random behavior frequently, in which it will simply not respond to any
 request. I used curl to manually verify. It appears that the sever is able
 to accept connection, but it doesn't seem to accept any data as I cannot
 see any logs on the server side.

 The version we are using is mina-core 2.0.0-M6.

 Right now I am running out of ideas to debug this issue. So, can anyone
 offer some advice about what I can try to investigate this issue? If you
 need more details to understand the issue, please, please let me know. I
 will do my best to provide them.

 Many thanks.

 Best regards,
 Jim

Re: Session Sharing between applications

[Jon V.]

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 the received data
into the other IoSession.
On Mar 21, 2015 8:15 PM, Emmanuel Lécharny elecha...@gmail.com wrote:

 Le 21/03/15 17:33, Nikunj Patel a écrit :
  Dear All,
 
  I am going to develop network application using apache mina where
  application continually connected to remote server and also open port for
  client connection. When any client send message it will pass to remote
  server and response received from remote server it will pass to
 respective
  client which request has sent. My TCP Server must keep live connection
 with
  remote tcp server.
 
  architecture would be like
 
  TCP Client1 -- TCP Server -- Remote TCP Server
  TCP Client2
  .
  .
  TCP Clientn
 
  I have no idea how I can response back to tcp client when response
 received
  from remote tcp server.
 
  Please find attached my sample application for your reference and guide
 me
  how can I achieve it.
 
 The attachement didn't make it.

 Regardless, you are just writing a proxy server. I suppose that your
 proxy send a request to the remote server when it receives a message
 from your client, so sending back a response to this client when your
 server receives the response from the remote server is as simple as
 writing the response into the client session.

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

[Jon V.]

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 caller of write()
On Aug 16, 2014 2:45 PM, Jens Reimann ct...@dentrassi.de wrote:

 Yes, but the call to write() can originate from any thread and thus is not
 thread safe. All handlers in the write call get called unsynchronized. This
 also causes issues at other locations like the gzip filter.

 There are two options, synchronize write calls yourself by surrounding
 calls to write() with a mutex lock or semaphore.

 Or, if you cannot synchronize all write calls (e.g. when you use the
 Heartbeat filter, which calls write on its own) you need to use the
 ExecutorFilter (I think that was the name). This can be used to force read
 and write calls to a specific thread.

 Jens

 On Aug 16, 2014 8:18 PM, Emmanuel Lecharny elecha...@apache.org wrote:
 
  From the top of my head (I'm in a bus atm), each session has its own
 chain
  of filter, thus its own instance. The variable is not shared, and the
  session is always handled by the same IoPricessor, so the same thread.
 That
  should be thread safe unless you start doing weird things with an
 executor.
  Le 16 août 2014 18:06, Kai ZHANG longb...@gmail.com a écrit :
 
   Hi,
  
   I am a beginer of Mina. I read the api doc located at:
  
   /http://mina.apache.org/mina-project/apidocs/org/apache/
   mina/core/session/IoSession.html/
  
   It says that IoSession is thread-safe.
  
   But when I read the source of mina trunk branch. I found the
   IoSession.write() method may be not thread-safe. The java source file
 is :
  
   /./core/src/main/java/org/apache/mina/session/AbstractIoSession.java/
  
   Here is the method calling chain:
   AbstractIoSession.write()
   - AbstractIoSession.doWriteWithFuture
   - AbstractIoSession.processMessageWriting()
   - AbstractIoFilter.messageWriting()
   - AbstractIoSession.callWriteNextFilter()
  
   The code which is thread-safe reside in
 AbstractIoSession.callWriteNextFilter(),
   here is the code:
  
  /**
* process session message received event using the filter
  chain. To be called by the session {@link SelectorLoop} .
*
* @param message the received message
*/
   @Override
   public void callWriteNextFilter(WriteRequest message) {
   if (IS_DEBUG) {
   LOG.debug(calling next filter for writing for message
  '{}' position : {}, message, writeChainPosition);
   }
  
  /writeChainPosition--;/
  
   if (writeChainPosition  0 || chain.length == 0) {
   // end of chain processing
   enqueueWriteRequest(message);
   } else {
   chain[writeChainPosition].messageWriting(this, message,
  this);
   }
  
  /writeChainPosition++;/
   }
  
   Here the variable writeChainPosition is not thread-safe, If more than
   one thread call IoSession.write() concurrently, the
 writeChainPosition
   may have race condition.
  
   The result is some of the IoFilter may be skipped or called twice, and
 the
   message data passed down the filter chain may be broken.
  
   Could you tell me if my understanding is correct?
  
   Is IoSession.write() method designed to be thread-safe or should I use
 a
   lock for every concurrent IoSession.write() operation?
  
  
  
  
  

Re: Performance issue - NioSocketAcceptor

[Jon V.]

Yes, if the iohandle is blocking the execution then that would obviously do
it   That said I am not sure how the reactor threading in Mina would handle
very small messages sent in a way that prevents the data from aggregating.
Difference is 20k individual triggers or 1k triggers for the same 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-iperf-to-troubleshoot-speedthroughput-issues
 should do the trick. But here, with the client and the server o the same
 machine, I doubt the network could be the probelm.
 
  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.

 It's crystal clear that it's not a MINA problem per se. I'm able to
 saturate a 10Mb/s network with more than 20 000 messages per second on
 my computer on my LDAP server using MINA.

 One of the possible reasons : the server IoHandler is doing weird things
 (like waiting for an external resource before releasing the thread).
 Typical use case : sendinhg the received data to a remote data base.

 So bottom line : what do you do on the server side with the data you
 receive ?

Re: Performance issue - NioSocketAcceptor

[Jon V.]

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.
Hi there,

I am using Apache Mina to develop a TCP server that handles incoming XML
messages. There is around 60 to 80 concurrent messages sending to TCP
server per minute. If the messages below 60 the TCP server is working fine
and there is no delay receiving the request but problem arises when more
than 60.

 The problem is the messages didn't reach/receive on time. From the log
file I can see the messages are accumulated some where and after a while
(between 10 to 30 seconds) all accumulated messages release to TCP server.
Messages reach server on time is very critical in my application.

I try to fine tune the NioSocketAcceptor number of thread, increase read
buffer size, increase io buffer size but all do not help.

I executed a stress test on my local machine using Jmeter with same
throughput and the result is still negative. When NioSocketAcceptor starts
accumulating messages I can see Jmeter is throwing:
jmeter.protocol.tcp.sampler.TCPSampler:  java.net.SocketException: Software
caused connection abort: socket write error.
To me this look like NioSocketAcceptor is blocking the messages. After a
while it back to normal again.

Please take note the JMeter and TCP Server are running on the same machine
 - this is to isolate network factor when doing testing. From the result,
seem like it is not something related to network so now I am just focus on
fine tuning the NioSocketAcceptor.

Below is my codes. I will be grateful if anyone can tell me what is going
on internally. I have been stuck in this problem for few months but still
can't find any solution.

IoAcceptor acceptor = new NioSocketAcceptor();

 // Handle the incoming XML message
 PMMessageHandler pmMessageHandler = new PMMessageHandler();

// Initialise Spring configuration
PMUtils pmUtils = PMUtils.getInstance();

pmMessageHandler.setDataProcessor((DataProcessor)pmUtils.getBean(dataProcessor));

//pmMessageHandler.setPmManager((PMManager)pmUtils.getBean(pmManager));
acceptor.setHandler(  pmMessageHandler );

// Set the filter chain for log and translate incoming message to char
 acceptor.getFilterChain().addLast( logger, new
LoggingFilter() );
TextLineCodecFactory textLineCodecFactory = new
TextLineCodecFactory(Charset.forName( UTF-8 ));

textLineCodecFactory.setDecoderMaxLineLength(PMServer.maxLineLength);
acceptor.getFilterChain().addLast( codec, new
ProtocolCodecFilter( textLineCodecFactory ));

// Configure buffer size
IoBuffer buffer = IoBuffer.allocate(5120, false);
buffer.setAutoExpand(true);



try {

 // Configure a port for TCP listener
acceptor.bind( new InetSocketAddress(PMServer.port));

Thanks,
KJ

Re: Help with so many concurrent connections

[Jon V.]

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 database driver works.
On Jul 15, 2014 5:11 PM, Eray Yuksek eyuksek2...@yahoo.com.invalid
wrote:

 Hi,

 I am new in the list so I am not sure if I am sending my question to the
 right list since I know that there is another list called
 d...@mina.apache.org.
 I apologize in advance if this is not the right list and in that case,
 please let me know.

 Here is our problem:

 We try to handle thousands of concurrent connections through Mina, and
 what we do is that we put all of the messages goes through Mina
 successfully into a buffer with size 10K and then one by one we get the
 messages from the buffer and insert into single sql server table. However
 under heavy load, our inserts into the database table slows down
 significantly. We actually had no buffer before but at that time, it was
 even worse to insert all those concurrent messages right away into the
 table. We thought buffer would give us time to store the data under heavy
 load and we could get the messages one by one from the buffer and insert
 into the DB. But nowadays, even buffer is not enough for us.
 we are not sure whether the buffer is holding us back or the DB inserts
 but somehow we have to find a way to insert all the concurrent messages
 into that one DB table.
 (the buffer we are using it BoundedFifoBuffer and we use it as
 synchronized buffer)


 Thanks for your help in advance.

 Eray Yuksek


 E.Y.

Re: sshd - async interface

[Jon V.]

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. sybersn...@gmail.com wrote:
 
  NIO controls and deals with the selectors. Async IO is a part of that but
  is not the same thing. Async io means that if a write cannot be fully
  flushed. It will not block until it can be. NIO provides us the events to
  tell us that data is available in the socket.

 Async IO is the ability for a single thread to perform (multiplex) IO
 (connect, read, write, close etc..) for multiple file descriptors.

 As far as I know, without NIO you cannot achieve that in Java.

 There is no sense in read or write without blocking if you cannot wait
 (vs actively poll) for an event.

  On Apr 19, 2014 4:56 AM, Alon Bar-Lev alon.bar...@gmail.com wrote:
 
   On Sat, Apr 19, 2014 at 10:58 AM, Emmanuel Lécharny 
 elecha...@gmail.com
   wrote:
Le 4/19/14 9:45 AM, Alon Bar-Lev a écrit :
On Sat, Apr 19, 2014 at 10:38 AM, Emmanuel Lécharny 
   elecha...@gmail.com wrote:
Le 4/19/14 9:13 AM, Alon Bar-Lev a écrit :
Hi,
   
The mission of async is to avoid having threads at all, or at
 least
   O(1).
   
As you have underline internal/private low level channels for
 socket
processing, and public high level channels to communicate with
application, there should be a mechanism for library to request
 wake
up for these low level channels.
   
Another option is to avoid using sockets at all within the
implementation and require application to manage the sockets and
 pipe
socket data into the library.
   
I understand this is conceptional change than what we have now,
 but
this what will enable scale without abusing system threads or have
nondeterministic behaviour in high load.
There are a few important things you have to know about async and
   threads :
- the extra cost for dealing with async connection is around 30%.
 That
all but free
- a standard system can easily deal with a few thousands of threads
   
Now, unless you define what is high load, I don't really see what
   kind
of advantage we can get with an async implementation.
   
FTR, when MINA was initially created, it was because there was a
 need
for a system supporting potentially ten of thousands of
 connections. Is
that what you are targetting ?
Yes, using work threads that are derived per # of CPUs, no more.
I am far from the pure Java world... but if async IO is 30%
insufficient, maybe it worth to use libssh (C) and communicate with
 it
using single socket from java, delegating IO outside of java.
IO are already delegated outside on Java. Eveything IO related is
written in C and wrapped into Java class.
   
The extra cost when using NIO is due to the management of SelectorKey
lists (with the various steps involved when dealing with those
 lists).
   
All in all, when it comes to process IO, Java does not really add
 some
extra cost over a plain C implementation. It's not the same story
 when
using NIO, especially when dealing with many concurrent connections.
   
  
   So I am confused... Java does not add cost to async IO, but NIO does?
   While NIO is the only interface to Java async IO?
  

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

[Jon V.]

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 regarding missing packets 39-55.  Those are packets
that probably exist in the native io buffer. No application can read what
is in there? The assumption is that data in there will be written. However
a socket can be closed while data exists in this buffer.
Your higher level proto needs to send back an ACK with a packet number. You
need to handle this yourself.
On Mar 31, 2014 12:22 PM, TRAN Antoine antoine-b.t...@thalesgroup.com
wrote:

 Hi,

 Goal
 I would like to improve the reliability of our program using Mina.
 From the research and tests I have made, it seems:

  1.  TCP reliability is not enough for messages not to be lost.
  2.  Mina does not offer message sent acknowledgment.
  3.  Lost of connection can be detected with the IOException that Mina
 will throw in IoHandlerAdapter#exceptionCaught.
  4.  If the remote socket is closed when writing to it, Mina will throw an
 WriteToClosedSessionException in IoHandlerAdapter#exceptionCaught.

 Based on these statements, I have made a program that will use a buffer in
 case Mina cannot send messages. When the reconnection is done, it will
 flush the buffer and then resend as normal.

 Problems
 This is my test. I send 100 messages at 1/sec. I removed the ethernet
 cable after the 39th messages, put it back a few seconds later. As a
 consequence, I lost some messages. Here are some elements that might help.

  1.  The written data are from messages 1 to 39 and from 55 to 100 (so
 some messages are lost).
  2.  The Mina server and client notice an IOException.
  3.  The Mina client gets the IOException way too late (after the 53th
 message).
  4.  The WriteToClosedSessionException that is supposed to contain all the
 messages that could not be sent contains only the messages from 55 to 58
 (expected from 40 to 58).

 Questions

  1.  Are the above statements correct?
  2.  Is my workaround a good solution or does Mina already offer a better
 way of keeping messages before sending them in case of network error?

 Thank you.

 Yours sincerely,

 Antoine

Re: How to write byte[] to an IoSession

[Jon V.]

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 = in.array();
 int length = inBytes[0];
 inBytes = Arrays.copyOfRange(inBytes, 1, length + 1);
 ByteString incomingMessage = ByteString.copyFrom(inBytes);

 But when I try to do

 someIoSession.write(incomingMessage.toByteArray());

 I get the following error.

 Don't know how to handle message of type .  Are you missing a protocol
 encoder?

 How can i just write the bytes into an IoSession ??




 --
 View this message in context:
 http://apache-mina.10907.n7.nabble.com/How-to-write-byte-to-an-IoSession-tp41542.html
 Sent from the Apache MINA User Forum mailing list archive at Nabble.com.

Re: Separate Threads for sending and receiving

[Jon V.]

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 write out the response
 to session.write. But it is not immediately written out. During my
 process of the message additional message are received and process
 before my response message is delivered.


 It all depends on the session processing the incoming message. This is a
 multi-threaded framework, messages comming from dfferent sessions may be
 processed in //.


 OK because, i implement the client i only have one session, thus this
 doesn't complicate my issue.

  Now i figured out that only one thread handles read/write request.

 Absolutely.

  And
 the read request is processed before any write request?


 depends. If the socket is full when the thread is trying to write into
 it, then it give up, and a new incoming message might be processed
 before the write can be done. Eventually, a lot of messages can be
 stacked waiting for being written.


  So how can i
 either have a separate Thread for sending and receiving.


 You can use an Executor in the chain, even two executors if needed (one
 for reads, one for writes). But at the nend of the day, it does not
 somlve your issue.

  Or alternative
 how to tell the NioSocketprocess to take priority processing of write
 requests?


 You can idle the read while the write is not processed. The way it works
 woudl be something like :

 - process a read (in the handler)
 - idle the read (session.suspendRead())
 - write the response

 Then in the handler, react on the messageSent event :

 - process the messageSent event
 - un-idle the read (session.resumeRead())

 You now can process a new read.

 Think twice before implementing such a logic though !!! I would question
 the fact that the client is sending messages but not reading the
 response fast enough...


 You correct, your purposed solution doesn't solve my issue. At least
 i tried the ExecutorFilter, the suspendRead solution feels already from
 the reading ugly, so i didn't give it try. And because i'm the client
 i think i have a performance problem. But my investigation shows me
 that from receiving the command in my IoHandler to writing the response
 with session.write i need around 2 ms. But than it takes about 70 ms to
 get a logging about form the NioProcessor saying it send the message.

 Normally the NioProcess will not received any additional commands from
 between the first and sending my response. But under heavy load it will
 receive additional commands before sending my response and thus delay
 it additional about some ms.

 My goal is to get this 70ms as low as possible, and you say the cause
 for the long processing time is that the socket is full? Can a socket be
 full because of message collisions on the wire?

 Oh by the way i know i can get faster, because in raw cases the 70ms
 drop down to 2ms. Thus the whole command response cycle is processed
 in about 4ms.

 best regards

 Max

 --
 _

 http://facilityboss.biz
 http://facebook.com/facilityboss
 _

 Max Larsson
 facilityboss GmbH
 Rheinstrasse 75, 64295 Darmstadt / Germany
 Handelsregister Darmstadt, HRB 86193
 Geschäftsführer: Dipl.-Inform. Max Lars Robert Larsson

 Mobil: +49 179 2184428
 Email: max.lars...@facilityboss.biz

Re: handling a slow client

[Jon V.]

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 consider, as some script kiddy tools for
 attacking websites behave this way.


 On Fri, Nov 22, 2013 at 4:00 PM, Binole, Bill 
 william.j.bin...@questdiagnostics.com wrote:

  I was wondering how to handle a slow client.  So say my client writes 100
  bytes and then 5 seconds later writes another 100 bytes.  What is being
  sent is delimited with a start and end block that I am checking for to
  insure I have a full message.  I have written a decoder based on the
  CumulativeProtocolDecoder that works just fine except for the case of a
  slow client.  In this case the buffer is drained before the end block is
  received and it is causing a bogus error condition.  Is there a way to
  handle this within the decoder?
 
  Bill
 
  __
  The contents of this message, together with any attachments, are intended
  only for the use of the person(s) to which they are addressed and may
  contain confidential and/or privileged information. Further, any medical
  information herein is confidential and protected by law. It is unlawful
 for
  unauthorized persons to use, review, copy, disclose, or disseminate
  confidential medical information. If you are not the intended recipient,
  immediately advise the sender and delete this message and any
 attachments.
  Any distribution, or copying of this message, or any attachment, is
  prohibited.




 --
 Cheers,

 Andrew Pennebaker
 apenneba...@42six.com

Re: Can MINA create proxy servers?

[Jon V.]

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?

[Jon V.]

I imagine you could build a proxy in Mina. However I don't believe any
prebuilt proxies exist. May be easier to just use nginx.
On Nov 22, 2013 3:12 PM, Andrew Pennebaker apenneba...@42six.com wrote:

 I ask, because I'm using Mina to spin up little test HTTP/FTP servers, to
 test a web crawler. 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 at.
  On Nov 22, 2013 11:59 AM, Andrew Pennebaker apenneba...@42six.com
  wrote:
 
   That would be cool!
  
   --
   Cheers,
  
   Andrew Pennebaker
   apenneba...@42six.com
  
 



 --
 Cheers,

 Andrew Pennebaker
 apenneba...@42six.com

Re: Can MINA create proxy servers?

[Jon V.]

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 be easier to just use nginx.


 http://mina.apache.org/mina-project/xref-test/org/apache/mina/example/proxy/ProxyTestClient.html

 and


 http://mina.apache.org/mina-project/xref/org/apache/mina/proxy/ProxyConnector.html

 It's included in the Mina core lib.


 --
 Regards,
 Cordialement,
 Emmanuel Lécharny
 www.iktek.com

Re: IoSession write method thread safe?

[Jon V.]

Synchronized is important to update the thread caches. More so when
different sessions are bound to different threads.
On Nov 21, 2013 10:45 AM, Alexandros Touloupis atoulou...@gmail.com
wrote:

 Thanks for the reply!

 Yes I agree that synchronizing to the whole IoSession instance each time
 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 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 sockets.
  On Nov 21, 2013 10:23 AM, Alexandros Touloupis atoulou...@gmail.com
  wrote:
 
   Hello all,
  
   I am using mina 2.0.8 since I needed a patch included in this release.
  
   I face the following problem:
  
   The concept of the application is to multiplex inbound to outbound
   connections in MINA. This means that there might be 2 incoming
  connections
   that are multiplexed in one outbound connection. During my tests i
  randomly
   see this exception:
  
   java.nio.InvalidMarkException
   at java.nio.Buffer.reset(Buffer.java:298) ~[?:1.7.0_40]
   at
  
  
 
 org.apache.mina.core.buffer.AbstractIoBuffer.reset(AbstractIoBuffer.java:414)
   ~[mina-core-2.0.8-PATCHED.jar:?]
   at
  
  
 
 org.apache.mina.core.polling.AbstractPollingIoProcessor.writeBuffer(AbstractPollingIoProcessor.java:926)
   ~[mina-core-2.0.8-SNAPSHOT.jar:?]
   at
  
  
 
 org.apache.mina.core.polling.AbstractPollingIoProcessor.flushNow(AbstractPollingIoProcessor.java:844)
   [mina-core-2.0.8-SNAPSHOT.jar:?]
   at
  
  
 
 org.apache.mina.core.polling.AbstractPollingIoProcessor.flush(AbstractPollingIoProcessor.java:772)
   [mina-core-2.0.8-SNAPSHOT.jar:?]
   at
  
  
 
 org.apache.mina.core.polling.AbstractPollingIoProcessor.access$700(AbstractPollingIoProcessor.java:67)
   [mina-core-2.0.8-SNAPSHOT.jar:?]
   at
  
  
 
 org.apache.mina.core.polling.AbstractPollingIoProcessor$Processor.run(AbstractPollingIoProcessor.java:1112)
   [mina-core-2.0.8-SNAPSHOT.jar:?]
   at
  
  
 
 org.apache.mina.util.NamePreservingRunnable.run(NamePreservingRunnable.java:64)
   [mina-core-2.0.8-SNAPSHOT.jar:?]
   at
  
  
 
 java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
   [?:1.7.0_40]
   at
  
  
 
 java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
   [?:1.7.0_40]
   at java.lang.Thread.run(Thread.java:724) [?:1.7.0_40]
  
   After some investigation i noticed that this happened when 2 threads
   decided to write to the same IoSession a message the same time. My
   assumption is that the #IoSession.write method is not thread safe,
 thus i
   synchronized the session instance before executing the write method in
 my
   handlers and up until now i haven't faced the same issue. Is my
  assumption
   valid?
  
   Thank you in advance,
   Alex
  
 

Re: IoSession write method thread safe?

[Jon V.]

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 sockets.
On Nov 21, 2013 10:23 AM, Alexandros Touloupis atoulou...@gmail.com
wrote:

 Hello all,

 I am using mina 2.0.8 since I needed a patch included in this release.

 I face the following problem:

 The concept of the application is to multiplex inbound to outbound
 connections in MINA. This means that there might be 2 incoming connections
 that are multiplexed in one outbound connection. During my tests i randomly
 see this exception:

 java.nio.InvalidMarkException
 at java.nio.Buffer.reset(Buffer.java:298) ~[?:1.7.0_40]
 at

 org.apache.mina.core.buffer.AbstractIoBuffer.reset(AbstractIoBuffer.java:414)
 ~[mina-core-2.0.8-PATCHED.jar:?]
 at

 org.apache.mina.core.polling.AbstractPollingIoProcessor.writeBuffer(AbstractPollingIoProcessor.java:926)
 ~[mina-core-2.0.8-SNAPSHOT.jar:?]
 at

 org.apache.mina.core.polling.AbstractPollingIoProcessor.flushNow(AbstractPollingIoProcessor.java:844)
 [mina-core-2.0.8-SNAPSHOT.jar:?]
 at

 org.apache.mina.core.polling.AbstractPollingIoProcessor.flush(AbstractPollingIoProcessor.java:772)
 [mina-core-2.0.8-SNAPSHOT.jar:?]
 at

 org.apache.mina.core.polling.AbstractPollingIoProcessor.access$700(AbstractPollingIoProcessor.java:67)
 [mina-core-2.0.8-SNAPSHOT.jar:?]
 at

 org.apache.mina.core.polling.AbstractPollingIoProcessor$Processor.run(AbstractPollingIoProcessor.java:1112)
 [mina-core-2.0.8-SNAPSHOT.jar:?]
 at

 org.apache.mina.util.NamePreservingRunnable.run(NamePreservingRunnable.java:64)
 [mina-core-2.0.8-SNAPSHOT.jar:?]
 at

 java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
 [?:1.7.0_40]
 at

 java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
 [?:1.7.0_40]
 at java.lang.Thread.run(Thread.java:724) [?:1.7.0_40]

 After some investigation i noticed that this happened when 2 threads
 decided to write to the same IoSession a message the same time. My
 assumption is that the #IoSession.write method is not thread safe, thus i
 synchronized the session instance before executing the write method in my
 handlers and up until now i haven't faced the same issue. Is my assumption
 valid?

 Thank you in advance,
 Alex

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

[Jon V.]

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 called.  If this buffer overflows the TCP connections drop in a bad way.


On Mon, Sep 30, 2013 at 11:40 AM, praveen developsfuturet...@gmail.comwrote:

 i have set the backlog to 12 but still i have the same problem



 --
 View this message in context:
 http://apache-mina.10907.n7.nabble.com/Unable-to-open-more-than-1000-ports-in-Mina-Client-tp39936p39962.html
 Sent from the Apache MINA User Forum mailing list archive at Nabble.com.

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

[Jon V.]

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 expensive in Windows. (maybe because
they are not totally independent of each other).  Where I can use 8 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 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 called.  If this buffer overflows the TCP connections drop in a bad
 way.

 As I already said in a previous mail, this 1000 port limit is extremelly
 weird.

 In any case, this has nothing to do with MINA and any setting on MINA.

 Chck :
 - your firewall
 - your router setting.
 - consider using a decent OS instead of Windows 2008 which is, all in
 all, just 6 years old...


 --
 Regards,
 Cordialement,
 Emmanuel Lécharny
 www.iktek.com

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

[Jon V.]

.But when 1000 clients connect simultaneously more than
50% of the connections will not respond with in 3 seconds and so the
connection is closed by the client.

So it sounds to me like he was connecting all of them at the same time.
 Hence my backlog statement.  If he connected them 5-10 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 expensive in Windows. (maybe because
 they are not totally independent of each other).  Where I can use 8 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 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 called.  If this buffer overflows the TCP connections drop in a bad
 way.

 As I already said in a previous mail, this 1000 port limit is extremelly
 weird.

 In any case, this has nothing to do with MINA and any setting on MINA.

 Chck :
 - your firewall
 - your router setting.
 - consider using a decent OS instead of Windows 2008 which is, all in
 all, just 6 years old...


 --
 Regards,
 Cordialement,
 Emmanuel Lécharny
 www.iktek.com

Re: Multithreaded UDP server vs. OrderedThreadPoolExecutor

[Jon V.]

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

 On Tue, 24 Sep 2013 11:51:04 -0700
 Emmanuel Lécharny elecha...@gmail.com wrote:

  Le 9/24/13 9:47 AM, Christian Hammers a écrit :
   Hello
  
   I've written a server for a proprietary UDP protocol that needs to
 answer with
   exactly one packet for every incoming packet. All packets are
 independent from
   each other so that I like to have each one handled asynchronously by a
   different thread from a pool with a certain maximum size.
  
   So far I used this:
  
   FooCodec codec = new FooCodec(); // encoder and decoder in one
 class
   InetSocketAddress local = new InetSocketAddress(port);
  
   acceptor = new NioDatagramAcceptor();
   acceptor.getFilterChain().addLast(executor, new
 ExecutorFilter(maxThreads));
   acceptor.getFilterChain().addLast(codec, new
 ProtocolCodecFilter(codec, codec));
   acceptor.setHandler(handler);
   acceptor.bind(local);
  
   At first it seemed that each incoming UDP package, due to its
 stateless nature,
   creates a different Mina session and each Mina session runs in its
 own
   thread so that the default ExecutorFilter implementation
   OrderedThreadPoolExecutor is fine as I certainly don't want the
 session closed
   if there is still a response before it in the queue.
 
  That's not what is currently happening. MINA creates a session for each
  IP address, so if all your UDP messages are issued by one single machine
  using one IP address, then all the UDP messages will be processed by the
  same thread.

 Ah! They are indeed all coming from the same IP+Port.
 (I used to do a session.write(message); session.close() in my
 messageReceived()
 method so it seemed to be different sessions but that was probably a wrong
 approach as well)


  Adding an executor in the chain just make it likely that a separate
  thread will process each UDP message, but the session remains shared.
 
  That explains what you see.

 OK

  Just get rid of the executor, and increase the number of IoProcessor,
  you will be able to spread the load on many threads, and each new UDP
  message will be completely processed before the next UDP message can
  start to be processed.

 How can I increase the number of IoProcessor? Unlike the
 NioSocketAcceptor, the
 NioDatagramAcceptor constructor has no processorCount argumment.

 But regarding your last half sentence: Do I understand correctly that each
 IoProcessor
 thread processes one message after the other but as I can have several of
 the
 IoProcessor threads, the whole application will still be able to handle
 incoming request asynchronously?

 Best Regards,

 -christian-

Re: Multithreaded UDP server vs. OrderedThreadPoolExecutor

[Jon V.]

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 one packet for every incoming packet. All packets are independent
 from
 each other so that I like to have each one handled asynchronously by a
 different thread from a pool with a certain maximum size.

 So far I used this:

 FooCodec codec = new FooCodec(); // encoder and decoder in one
 class
 InetSocketAddress local = new InetSocketAddress(port);

 acceptor = new NioDatagramAcceptor();
 acceptor.getFilterChain().addLast(executor, new
 ExecutorFilter(maxThreads));
 acceptor.getFilterChain().addLast(codec, new
 ProtocolCodecFilter(codec, codec));
 acceptor.setHandler(handler);
 acceptor.bind(local);

 At first it seemed that each incoming UDP package, due to its stateless
 nature,
 creates a different Mina session and each Mina session runs in its own
 thread so that the default ExecutorFilter implementation
 OrderedThreadPoolExecutor is fine as I certainly don't want the session
 closed
 if there is still a response before it in the queue.

 Now under a bit of load, one session was a bit slower and hanging for 1,5s
 and
 it suddenly happened that multiple UDP packets were put into the same
 Mina session which lead to the following log messages:

   2013-09-24 10:56:02,910 DEBUG
 org.apache.mina.filter.executor.OrderedThreadPoolExecutor   :
 Adding event MESSAGE_RECEIVED to session 17529
 Queue : [MESSAGE_RECEIVED, , MESSAGE_RECEIVED, , MESSAGE_RECEIVED ... ]
 (OrderedThreadPoolExecutor.java:431)

 This now completely destroys the asynchronous behaviour as all packeges
 had to wait
 for the first one in the queue even if there were several executor threads
 idling
 around!

 How should I fix this problem? Using an UnorderedThreadPoolExecutor?
 Or setting acceptor.getSessionConfig().setIdleTime(IdleStatus.BOTH_IDLE,
 0);
 Or by somehow else force the creation of a Mina session for each UDP
 packet?

 Best Regards

 -christian-


 P.S.: http://mina.apache.org/mina-project/documentation.html contains a
   links to a UDP Tutorial which sadly refers to a page that does
   only have APR and Serial tutorials.

Re: Trouble with messageReceived()

[Jon V.]

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 tell me what I am doing
 wrong. I don't seem to be getting the messageReceived() called on either
 the client or the server code. I have been able to get sumup example to run
 and I used code from that example to mashup this test code. I am rung OS/X
 with java 7 using mina 2.0.7. The logger shows the CREATED, OPENNED, and
 CLOSED events for the session but the messageReceived() routine never seems
 to be called. Is there something simple I am missing here?

 Thanks,
 Mike Karrys
 mike.kar...@ngc.com

 Server Code:

 public class Main extends IoHandlerAdapter {

 private static final int SERVER_PORT = 8080;
 private final static Logger LOGGER =
 LoggerFactory.getLogger(org.jeuron.test.mina.fl.server.Main.class);

 public void init() throws IOException {
 NioSocketAcceptor acceptor = new NioSocketAcceptor();

 // Prepare the service configuration.
 acceptor.getFilterChain().addLast(
 codec,
 new ProtocolCodecFilter(
 new ObjectSerializationCodecFactory()));

 acceptor.getFilterChain().addLast(logger, new LoggingFilter());

 acceptor.setHandler(this);
 acceptor.bind(new InetSocketAddress(SERVER_PORT));

 System.out.println(Listening on port  + SERVER_PORT);

 }

 public Map process(Map record) {
 int count = (Integer) record.get(Field.COUNT);
 int length = (Integer) record.get(Field.LENGTH);
 String input = (String) record.get(Field.CONTENTS);
 StringBuilder output = new StringBuilder();

 System.out.print(Processing( + (length + 1) + )\r);
 for (int i = input.length(); i  0; i--) {
 output.append(input.charAt(i - 1));
 }
 record.put(Field.CONTENTS, output.toString());

 if (count == length) {
 System.out.println(\nFinished Processing  + (count + 1));
 }

 return record;
 }

 @Override
 public void sessionOpened(IoSession session) {
 LOGGER.info(sessionOpened.);
 session.getConfig().setIdleTime(IdleStatus.BOTH_IDLE, 60);
 }

 @Override
 public void messageReceived(IoSession session, Object message) {
 Map input;
 Map output;

 LOGGER.info(messageReceived( + message + ));
 if (message instanceof Map) {
 input = (Map) message;
 output = process(input);
 session.write(output);
 }
 }

 @Override
 public void sessionIdle(IoSession session, IdleStatus status) {
 LOGGER.info(Disconnecting the idle.);
 session.close(true);
 }

 @Override
 public void exceptionCaught(IoSession session, Throwable cause) {
 session.close(true);
 }

 public static void main(String[] args) throws Exception {
 Main main = new Main();
 main.init();
 }
 }

 Client Code:

 public class Main extends IoHandlerAdapter {

 private final static Logger LOGGER =
 LoggerFactory.getLogger(org.jeuron.test.mina.fl.client.Main.class);
 private static final String HOSTNAME = localhost;
 private static final int PORT = 8080;
 private NioSocketConnector connector;
 private IoSession session;
 static char[] letter = {'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I',
 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X',
 'Y', 'Z'};
 private long startTime = 0;
 private long endTime = 0;
 private int maxSent = 0;
 private int receivedRecords = 0;
 private long totalBytes = 0;
 private int recordLengthValue = 0;
 private int recordCountValue = 0;

 public String readInput(String prompt) throws IOException {
 String input = null;

 System.out.println(prompt);
 BufferedReader br = new BufferedReader(new
 InputStreamReader(System.in));
 input = br.readLine();

 return input;
 }

 public boolean init() throws Exception {
 SocketAddress address = new InetSocketAddress(HOSTNAME, PORT);
 LOGGER.info(Main:init() address[ + address + ]);

 connector = new NioSocketConnector();
 connector.setHandler(this);
 for (;;) {
 try {
 ConnectFuture future = connector.connect(new
 InetSocketAddress(HOSTNAME, PORT));
 future.awaitUninterruptibly();
 session = future.getSession();
 break;
} catch (RuntimeIoException e) {
 System.err.println(Failed to connect.);
 e.printStackTrace();
 Thread.sleep(5000);
 }
 }

 //session = future1.getSession();

  

RE: EXT :Re: Trouble with messageReceived()

[Jon V.]

The ProtocolCodecFilter works on batches instead of stream based. You have
to be careful when you use it to ensure that you obtained all of the
processed messages. This could explain the dropped messages you are seeing.
On Sep 16, 2013 3:20 PM, Karrys, Michael (IS) mike.kar...@ngc.com wrote:

 Thanks. Knew it was something simple. I added the following and it now
 works.
connector.getFilterChain().addLast(
 codec,
 new ProtocolCodecFilter(
 new ObjectSerializationCodecFactory()));
 connector.getFilterChain().addLast(logger, new 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 PM
 To: users@mina.apache.org
 Subject: EXT :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 tell me what I am
  doing wrong. I don't seem to be getting the messageReceived() called
  on either the client or the server code. I have been able to get sumup
  example to run and I used code from that example to mashup this test
  code. I am rung OS/X with java 7 using mina 2.0.7. The logger shows
  the CREATED, OPENNED, and CLOSED events for the session but the
  messageReceived() routine never seems to be called. Is there something
 simple I am missing here?
 
  Thanks,
  Mike Karrys
  mike.kar...@ngc.com
 
  Server Code:
 
  public class Main extends IoHandlerAdapter {
 
  private static final int SERVER_PORT = 8080;
  private final static Logger LOGGER =
  LoggerFactory.getLogger(org.jeuron.test.mina.fl.server.Main.class);
 
  public void init() throws IOException {
  NioSocketAcceptor acceptor = new NioSocketAcceptor();
 
  // Prepare the service configuration.
  acceptor.getFilterChain().addLast(
  codec,
  new ProtocolCodecFilter(
  new ObjectSerializationCodecFactory()));
 
  acceptor.getFilterChain().addLast(logger, new
  LoggingFilter());
 
  acceptor.setHandler(this);
  acceptor.bind(new InetSocketAddress(SERVER_PORT));
 
  System.out.println(Listening on port  + SERVER_PORT);
 
  }
 
  public Map process(Map record) {
  int count = (Integer) record.get(Field.COUNT);
  int length = (Integer) record.get(Field.LENGTH);
  String input = (String) record.get(Field.CONTENTS);
  StringBuilder output = new StringBuilder();
 
  System.out.print(Processing( + (length + 1) + )\r);
  for (int i = input.length(); i  0; i--) {
  output.append(input.charAt(i - 1));
  }
  record.put(Field.CONTENTS, output.toString());
 
  if (count == length) {
  System.out.println(\nFinished Processing  + (count + 1));
  }
 
  return record;
  }
 
  @Override
  public void sessionOpened(IoSession session) {
  LOGGER.info(sessionOpened.);
  session.getConfig().setIdleTime(IdleStatus.BOTH_IDLE, 60);
  }
 
  @Override
  public void messageReceived(IoSession session, Object message) {
  Map input;
  Map output;
 
  LOGGER.info(messageReceived( + message + ));
  if (message instanceof Map) {
  input = (Map) message;
  output = process(input);
  session.write(output);
  }
  }
 
  @Override
  public void sessionIdle(IoSession session, IdleStatus status) {
  LOGGER.info(Disconnecting the idle.);
  session.close(true);
  }
 
  @Override
  public void exceptionCaught(IoSession session, Throwable cause) {
  session.close(true);
  }
 
  public static void main(String[] args) throws Exception {
  Main main = new Main();
  main.init();
  }
  }
 
  Client Code:
 
  public class Main extends IoHandlerAdapter {
 
  private final static Logger LOGGER =
  LoggerFactory.getLogger(org.jeuron.test.mina.fl.client.Main.class);
  private static final String HOSTNAME = localhost;
  private static final int PORT = 8080;
  private NioSocketConnector connector;
  private IoSession session;
  static char[] letter = {'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H',
  'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V',
  'W', 'X', 'Y', 'Z'};
  private long startTime = 0;
  private long endTime = 0;
  private int maxSent = 0;
  private int receivedRecords = 0;
  private long totalBytes = 0;
  private int recordLengthValue

Re: MINA support for WebSockets?

[Jon V.]

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 archives contain year-old exchanges with Dhruv
 Chopra of ShepHertz in which the ShepHertz company offers to donate
 WebSocket protocol filters to MINA.  I see that code in the developers'
 JIRA.  But that exchange seems to have ended after just a few messages, and
 it doesn't look like the donated code has subsequently been integrated into
 a release.  What are MINA's plans to support WebSockets?  I think it's
 clear that ability to add WebSockets IO to existing MINA-based server apps
 is attractive for many users.

 Thanks!

 --Mark

Re: Send/Receive IoBuffer

[Jon V.]

Change what I gave you to conform with MINA.  I simply wrote everything in
notepad to give you an idea.


On Tue, Sep 3, 2013 at 1:31 AM, Luciano Coelho luci...@lvcsistemas.comwrote:

 Thanks a lot for you answer, I will study about filters and try your
 solution too.

 Thanks too much really.

 -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 CompoundBufferFilter

 filter_received( ByteBuffer source, IoSession session, Filter next ) {
 ByteBuffer buffering = session.getAttribute( SOME_PROPERTY_BUFFER_NAME );

  if( buffering == null )
  {
 buffering = ByteBuffer.allocate( some size );  }

  buffering.put( source );
  buffering.flip();

  next.received( buffering, session );
 }

 This will compound buffer incoming data.

 logic_received( ByteBuffer source, IoSession session ) { while(
 source.hasRemaining() ) { source.mark(); if( source.remaining  4 ) {
 source.reset();
 return;
 }
  int size = source.getInt();
  if( source.remaining  size )
 {
 source.reset();
 return;
 }
  String data = new String( buffering.array, buffering.offset, size,
 ISO-8859-1); }  source.compact(); }

 Now you can simply return if you don't 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(16*1024), on
  connector as you told me.
 
  I changed now for what you answered me, like this:
 
  if(size  0){
  while (buffer.hasRemaining()){
  byte[] payload = new byte[size];
buffer.get(payload);
data = new String( payload, ISO-8859-1);
  }
  }
 
 
  Ok - Mina does not allow for compound rx_buffers.  This means you are
  going to have to manually buffer the message yourself.
 
 
  received( ByteBuffer source, IoSession session ) { while(
  source.hasRemaining() ) {  ByteBuffer buffering =
  session.getAttribute( SOME_PROPERTY_BUFFER_NAME );
 
  if( buffering != null )
  {
 // must be a new data
 int size = // read size somehow
 buffering = ByteBuffer.allocate( size );
 
 buffering.put(source);
 
 if( buffering.hasRemaining() )
 {
  session.setAttribute( SOME_PROPERTY_BUFFER_NAME, buffering );  return;
 }
 else
 {
   // message is complete
  session.clearAttribute( SOME_PROPERTY_BUFFER_NAME ); String data = new
  String( buffering.array, buffering.offset, buffering.limit,
  ISO-8859-1);
 }
  }
  else
  {
   buffering.put(source);
   if( !buffering.hasRemaining() )
   {
  // message is complete
  session.clearAttribute( SOME_PROPERTY_BUFFER_NAME );  String data =
  new String( buffering.array, buffering.offset, buffering.limit,
  ISO-8859-1); }  else { return; }  } } }
 
  In MINA you have to read everything out of that ByteBuffer or it will
  be lost forever.  This means you could get stuck only having 2 bytes
  in the buffer and not enough to read the size value.  So you may have
  to setup a secondary buffer to store partial size values if that scenario
 does occur.
 
 
 
  I'm receiving from de server 2685 bytes and I set the buffer to
  64*1024, and I'm still having the error: D/WEB(26934):
  java.nio.BufferUnderflowException.
 
  I saw that setting tcpnodelay = false becomes better, I got make 3
  requests without error, but after the error happens again, and I have
  to restart the session again.
  Do I have to finish something in the IoBuffer after receive the data?
  Note that I'm not opening another session, I open one time and stay
  changing data all the time. The rest is functioning very well, one by
  one minute I send something 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: 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 an indication of your Java knowledge.
  System.out is your access to stdout in Java.  It is what you use to
  print things to the console.  It has nothing to do with your
 application.
 
  Maybe posting on StackOverflow would be a good idea for general java
 help?
 
 
  
  
  
   Is that ok?
  
  
  
   I'm

Re: Send/Receive IoBuffer

[Jon V.]

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 an indication of your Java knowledge.  System.out is
your access to stdout in Java.  It is what you use to print things to the
console.  It has nothing to do with your application.

Maybe posting on StackOverflow would be a good idea for general java help?





 Is that ok?



 I'm receiving the answer in Android via GPRS like this:



 First I set this on sessionCreated

 session.getConfig().setReadBufferSize(16*1024);



You should set this in Mina config.  It is not good to change on the fly
like this.

Read the Quick Start Guide:
http://mina.apache.org/mina-project/quick-start-guide.html



 @Override public void messageReceived(IoSession session, Object message)
 throws Exception {

 IoBuffer  buffer
 = (IoBuffer)   message;

 int  service
 = (Integer) session.getAttribute(SERVICE);

 Listener   listener
 = (Listener)session.getAttribute(LISTENER);

 String   data
 = ;

 int  size
 = buffer.getInt();

 while (buffer.hasRemaining()){

data += (char) buffer.get();


Using the += is the worst possible way to build a String.

Since you are reading a String.  Are you sure that the String is encoded
using 8 bits?

byte[] payload = new byte[size];
buffer.get(payload);

String data = new String( payload, UTF-8);




 }

 if(size == data.trim().length()){


Why are you trying to trim the string?  Are you expecting whitespaces?



if(listener != null)
 listener.requisitionFinished(data, service);

 } else {

if(listener != null)
 listener.requisitionFinished(, service);

 }

 Log.d(WEB, SERVICE:  + service +  SIZE:  + size + 
 DATA:  + data);

 }



 Sometimes I receive the data in parts, when the message arriving is too
 long
 and it causes an error, even if the buffersize is bigger than the data I'm
 receiving.



 Thank's a lot.



 Luciano Coelho

Re: Send/Receive IoBuffer

[Jon V.]

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(16*1024), on connector as
 you told me.

 I changed now for what you answered me, like this:

 if(size  0){
 while (buffer.hasRemaining()){
 byte[] payload = new byte[size];
   buffer.get(payload);
   data = new String( payload, ISO-8859-1);
 }
 }


Ok - Mina does not allow for compound rx_buffers.  This means you are going
to have to manually buffer the message yourself.


received( ByteBuffer source, IoSession session )
{
while( source.hasRemaining() )
{
 ByteBuffer buffering = session.getAttribute( SOME_PROPERTY_BUFFER_NAME );

if( buffering != null )
{
   // must be a new data
   int size = // read size somehow
   buffering = ByteBuffer.allocate( size );

   buffering.put(source);

   if( buffering.hasRemaining() )
   {
session.setAttribute( SOME_PROPERTY_BUFFER_NAME, buffering );
 return;
   }
   else
   {
// message is complete
session.clearAttribute( SOME_PROPERTY_BUFFER_NAME );
String data = new String( buffering.array, buffering.offset,
buffering.limit,  ISO-8859-1);
   }
}
else
{
buffering.put(source);
 if( !buffering.hasRemaining() )
{
// message is complete
session.clearAttribute( SOME_PROPERTY_BUFFER_NAME );
String data = new String( buffering.array, buffering.offset,
buffering.limit,  ISO-8859-1);
}
else
{
return;
}
}
}
}

In MINA you have to read everything out of that ByteBuffer or it will be
lost forever.  This means you could get stuck only having 2 bytes in the
buffer and not enough to read the size value.  So you may have to setup a
secondary buffer to store partial size values if that scenario does occur.



 I'm receiving from de server 2685 bytes and I set the buffer to 64*1024,
 and
 I'm still having the error: D/WEB(26934):
 java.nio.BufferUnderflowException.

 I saw that setting tcpnodelay = false becomes better, I got make 3 requests
 without error, but after the error happens again, and I have to restart the
 session again.
 Do I have to finish something in the IoBuffer after receive the data? Note
 that I'm not opening another session, I open one time and stay changing
 data
 all the time. The rest is functioning very well, one by one minute I send
 something 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: 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 an indication of your Java knowledge.  System.out is
 your access to stdout in Java.  It is what you use to print things to the
 console.  It has nothing to do with your application.

 Maybe posting on StackOverflow would be a good idea for general java help?


 
 
 
  Is that ok?
 
 
 
  I'm receiving the answer in Android via GPRS like this:
 
 
 
  First I set this on sessionCreated
 
  session.getConfig().setReadBufferSize(16*1024);
 
 
 
 You should set this in Mina config.  It is not good to change on the fly
 like this.

 Read the Quick Start Guide:
 http://mina.apache.org/mina-project/quick-start-guide.html


 
  @Override public void messageReceived(IoSession session, Object
  message) throws Exception {
 
  IoBuffer  buffer
  = (IoBuffer)   message;
 
  int  service
  = (Integer) session.getAttribute(SERVICE);
 
  Listener   listener
  = (Listener)session.getAttribute(LISTENER);
 
  String   data
  = ;
 
  int  size
  = buffer.getInt();
 
  while (buffer.hasRemaining()){
 
 data += (char) buffer.get();
 

 Using the += is the worst possible way to build a String.

 Since you are reading a String.  Are you sure that the String is encoded
 using 8 bits?

 byte[] payload = new byte[size];
 buffer.get(payload);

 String data = new String( payload, UTF-8);



 
  }
 
  if(size == data.trim().length()){
 

 Why are you trying to trim the string?  Are you expecting whitespaces?


 
 if(listener != null)
  listener.requisitionFinished(data, service);
 
  } else

Re: Send/Receive IoBuffer

[Jon V.]

This is better:  I am not going to use the correct API but you can get the
idea.

Create a Filter called CompoundBufferFilter

filter_received( ByteBuffer source, IoSession session, Filter next )
{
 ByteBuffer buffering = session.getAttribute( SOME_PROPERTY_BUFFER_NAME );

 if( buffering == null )
 {
buffering = ByteBuffer.allocate( some size );
 }

 buffering.put( source );
 buffering.flip();

 next.received( buffering, session );
}

This will compound buffer incoming data.

logic_received( ByteBuffer source, IoSession session )
{
while( source.hasRemaining() )
{
source.mark();
if( source.remaining  4 )
{
source.reset();
return;
}
 int size = source.getInt();
 if( source.remaining  size )
{
source.reset();
return;
}
 String data = new String( buffering.array, buffering.offset, size,
 ISO-8859-1);
}
 source.compact();
}

Now you can simply return if you don't 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(16*1024), on connector as
 you told me.

 I changed now for what you answered me, like this:

 if(size  0){
 while (buffer.hasRemaining()){
 byte[] payload = new byte[size];
   buffer.get(payload);
   data = new String( payload, ISO-8859-1);
 }
 }


 Ok - Mina does not allow for compound rx_buffers.  This means you are
 going to have to manually buffer the message yourself.


 received( ByteBuffer source, IoSession session )
 {
 while( source.hasRemaining() )
 {
  ByteBuffer buffering = session.getAttribute( SOME_PROPERTY_BUFFER_NAME );

 if( buffering != null )
 {
// must be a new data
int size = // read size somehow
buffering = ByteBuffer.allocate( size );

buffering.put(source);

if( buffering.hasRemaining() )
{
 session.setAttribute( SOME_PROPERTY_BUFFER_NAME, buffering );
  return;
}
else
{
  // message is complete
 session.clearAttribute( SOME_PROPERTY_BUFFER_NAME );
 String data = new String( buffering.array, buffering.offset,
 buffering.limit,  ISO-8859-1);
}
 }
 else
 {
  buffering.put(source);
  if( !buffering.hasRemaining() )
  {
 // message is complete
 session.clearAttribute( SOME_PROPERTY_BUFFER_NAME );
  String data = new String( buffering.array, buffering.offset,
 buffering.limit,  ISO-8859-1);
 }
  else
 {
 return;
 }
  }
 }
 }

 In MINA you have to read everything out of that ByteBuffer or it will be
 lost forever.  This means you could get stuck only having 2 bytes in the
 buffer and not enough to read the size value.  So you may have to setup a
 secondary buffer to store partial size values if that scenario does occur.



 I'm receiving from de server 2685 bytes and I set the buffer to 64*1024,
 and
 I'm still having the error: D/WEB(26934):
 java.nio.BufferUnderflowException.

 I saw that setting tcpnodelay = false becomes better, I got make 3
 requests
 without error, but after the error happens again, and I have to restart
 the
 session again.
 Do I have to finish something in the IoBuffer after receive the data? Note
 that I'm not opening another session, I open one time and stay changing
 data
 all the time. The rest is functioning very well, one by one minute I send
 something 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: 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 an indication of your Java knowledge.  System.out is
 your access to stdout in Java.  It is what you use to print things to the
 console.  It has nothing to do with your application.

 Maybe posting on StackOverflow would be a good idea for general java help?


 
 
 
  Is that ok?
 
 
 
  I'm receiving the answer in Android via GPRS like this:
 
 
 
  First I set this on sessionCreated
 
  session.getConfig().setReadBufferSize(16*1024);
 
 
 
 You should set this in Mina config.  It is not good to change on the fly
 like this.

 Read the Quick Start Guide:
 http://mina.apache.org/mina-project/quick-start-guide.html


 
  @Override public void messageReceived(IoSession session, Object
  message) throws Exception {
 
  IoBuffer  buffer
  = (IoBuffer)   message;
 
  int  service

Re: Send/Receive IoBuffer

[Jon V.]

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.

 Create a Filter called CompoundBufferFilter

 filter_received( ByteBuffer source, IoSession session, Filter next )
 {
  ByteBuffer buffering = session.getAttribute( SOME_PROPERTY_BUFFER_NAME );

  if( buffering == null )
  {
 buffering = ByteBuffer.allocate( some size );
  }

  buffering.put( source );
  buffering.flip();

  next.received( buffering, session );
 }

 This will compound buffer incoming data.

 logic_received( ByteBuffer source, IoSession session )
 {
 while( source.hasRemaining() )
 {
  source.mark();
 if( source.remaining  4 )
 {
 source.reset();
 return;
 }
  int size = source.getInt();
  if( source.remaining  size )
  {
 source.reset();
 return;
 }
  String data = new String( buffering.array, buffering.offset, size,
  ISO-8859-1);
 }
  source.compact();
 }

 Now you can simply return if you don't 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(16*1024), on connector
 as
 you told me.

 I changed now for what you answered me, like this:

 if(size  0){
 while (buffer.hasRemaining()){
 byte[] payload = new byte[size];
   buffer.get(payload);
   data = new String( payload, ISO-8859-1);
 }
 }


 Ok - Mina does not allow for compound rx_buffers.  This means you are
 going to have to manually buffer the message yourself.


 received( ByteBuffer source, IoSession session )
 {
 while( source.hasRemaining() )
 {
  ByteBuffer buffering = session.getAttribute( SOME_PROPERTY_BUFFER_NAME );

 if( buffering != null )
 {
// must be a new data
int size = // read size somehow
buffering = ByteBuffer.allocate( size );

buffering.put(source);

if( buffering.hasRemaining() )
{
 session.setAttribute( SOME_PROPERTY_BUFFER_NAME, buffering );
  return;
}
else
{
  // message is complete
 session.clearAttribute( SOME_PROPERTY_BUFFER_NAME );
 String data = new String( buffering.array, buffering.offset,
 buffering.limit,  ISO-8859-1);
}
 }
 else
 {
  buffering.put(source);
  if( !buffering.hasRemaining() )
  {
 // message is complete
 session.clearAttribute( SOME_PROPERTY_BUFFER_NAME );
  String data = new String( buffering.array, buffering.offset,
 buffering.limit,  ISO-8859-1);
 }
  else
 {
 return;
 }
  }
 }
 }

 In MINA you have to read everything out of that ByteBuffer or it will be
 lost forever.  This means you could get stuck only having 2 bytes in the
 buffer and not enough to read the size value.  So you may have to setup a
 secondary buffer to store partial size values if that scenario does occur.



 I'm receiving from de server 2685 bytes and I set the buffer to 64*1024,
 and
 I'm still having the error: D/WEB(26934):
 java.nio.BufferUnderflowException.

 I saw that setting tcpnodelay = false becomes better, I got make 3
 requests
 without error, but after the error happens again, and I have to restart
 the
 session again.
 Do I have to finish something in the IoBuffer after receive the data?
 Note
 that I'm not opening another session, I open one time and stay changing
 data
 all the time. The rest is functioning very well, one by one minute I send
 something 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: 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 an indication of your Java knowledge.  System.out is
 your access to stdout in Java.  It is what you use to print things to the
 console.  It has nothing to do with your application.

 Maybe posting on StackOverflow would be a good idea for general java
 help?


 
 
 
  Is that ok?
 
 
 
  I'm receiving the answer in Android via GPRS like this:
 
 
 
  First I set this on sessionCreated
 
  session.getConfig().setReadBufferSize(16*1024);
 
 
 
 You should set this in Mina config.  It is not good to change on the fly
 like this.

 Read the Quick Start Guide:
 http://mina.apache.org/mina-project/quick-start-guide.html


 
  @Override public void

Re: Regarding DIRMINA-912

[Jon V.]

Sounds like you need a task queue and a single dispatcher thread. Write the
metered messages to the task queue and let it do the temporal management.
This can all be done with one thread.
On Aug 25, 2013 11:57 AM, Mike McKnight m...@mcknight4.com wrote:

 In my case I have units on the ground that need to have messages throttled
 when being sent to the units. In other words all of the queued write events
 going to each unit need to be metered out in an temporal fashion.  I have a
 separate thread pool available that can expand to the number of units on
 the ground, which is max around 50, and handles this metering process.
  Message processing for all but WRITE events needs to be unfettered in any
 way and can not be held up for writes being throttled.  So for sure I need
 another executor filter that handles all but WRITE events.

 I don't see how I can do this with a single executor filter. As events are
 processed in order in a queue and if I am waiting on a write then
 MESSAGE_RECEIVED events are not going to get processed. There are no
 separate SessionTasksQueue for the OrderedThreadPoolExecutor, just the one
 for all events waiting to be processed.

 Filter chain looks 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 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 McKnightm...@mcknight4.com  wrote:

  I am just curious if there is a downside to this issue that I am not
 seeing.  I have seen this in my MINA application; I have inbound/outbound
 executor filters as described in the issue and have seen where my inbound
 threads are *sometimes* processing IoEventType.WRITE events.  I don't see
 any adverse effects of this, my messages are being put on the wire, but
 am
 checking in with the list to be sure.

 I am running mina-2.0.7.

 https://issues.apache.org/jira/browse/DIRMINA-912https://issues.apache.org/**jira/browse/DIRMINA-912
 https:**//issues.apache.org/jira/**browse/DIRMINA-912https://issues.apache.org/jira/browse/DIRMINA-912
 

 Thanks!

 Mike

Re: Regarding DIRMINA-912

[Jon V.]

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 curious if there is a downside to this issue that I am not
 seeing.  I have seen this in my MINA application; I have inbound/outbound
 executor filters as described in the issue and have seen where my inbound
 threads are *sometimes* processing IoEventType.WRITE events.  I don't see
 any adverse effects of this, my messages are being put on the wire, but am
 checking in with the list to be sure.

 I am running mina-2.0.7.

 https://issues.apache.org/**jira/browse/DIRMINA-912https://issues.apache.org/jira/browse/DIRMINA-912

 Thanks!

 Mike

Re: AW: AW: NioSocketConnector/NioSocketAcceptor

[Jon V.]

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 happy to answer questions that involve things that are not
covered by the docs.
On Aug 15, 2013 11:31 AM, Emmanuel Lécharny elecha...@gmail.com wrote:

 Le 8/15/13 5:13 PM, Simo Chiegang, Boris Arthur RD-P8.1 a écrit :
  Ok that  means, when  I write new NioSocketConnector(), the number of
 threads to use isn't no limited. If I give a a processorCount (may be 2 ),
 I will limit the number of threads to 2. I'm wright?

 Yes.

 If you don't provide any value, the number of IoProcessor will be Nb
 Core + 1.

 usually, it's enough. If you need more, you can add an executor in the
 filter chain, and use a Pool. But keep in mind that whatever you do, if
 you are CPU bound, no matter how many threads you add, you are not going
 to have better throughput.

 --
 Regards,
 Cordialement,
 Emmanuel Lécharny
 www.iktek.com

Re: User authentication

[Jon V.]

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 mcmil...@gmail.com wrote:

 So that is actually the solution we had before, but the state machine was
 in an authentication filter, we had a lot of trouble transitioning from ok
 now I am done authenticating, I want to be in my IoHandler now

 Hunter
 On 8/7/2013 3:03 PM, Jon wrote:

 So just implement some sort if authentication using the state machine and
 a database record lookup.

 Sent from my iPhone

 On Aug 7, 2013, at 2:11 PM, Hunter McMillen mcmil...@gmail.com wrote:

  I misunderstood what you were asking, yes we have an abstract Command
 class that has derivatives for all of the commands that a user could enter.

 Hunter
 On 8/7/2013 1:13 PM, Jon wrote:

 You must be building some sort of communication format for it.

 Sent from my iPhone

 On Aug 7, 2013, at 1:06 PM, Hunter McMillen mcmil...@gmail.com wrote:

  We're not really implementing a protocol, we are just trying to make a
 text-based game that will seamlessly support hundreds to thousands of 
 users
 at a time.

 Hunter
 On 8/7/2013 12:02 PM, Jon wrote:

 This is more of a protocol implementation than a networking question.
  What kind of protocol are you implementing.

 Sent from my iPhone

 On Aug 7, 2013, at 11:34 AM, Hunter McMillen mcmil...@gmail.com
 wrote:

  We actually won't need anything as complex as LDAP, just a simple DB
 hash + salt lookup.

 Hunter
 On 8/7/2013 10:26 AM, Emmanuel Lécharny wrote:

 Le 8/7/13 4:11 PM, Hunter McMillen a écrit :

 What are the common ways to authenticate users using Mina?

 The first attempt I made was using a state machine, but I had
 problems
 integrating that with an IoHandler.

 In retrospect, the state machine seems like overkill; so I was
 hoping
 to get ideas for other ways to authenticate users, or maybe a link
 to
 an application that does some authentication.

 would it be terrible to do something like this?

 public void sessionCreated(IoSession session) {
  authenticateUser()  //  - Good idea? Bad Idea? Run in a
 separate thread?
 }

 Depends...

 let's see how it works with LDAP :

 - the user can connect on the server (and the sessionCreated event
 is
 handled), but the user will not be authentified at this point.

 - in order to authenticate the user, we need to know about the user.
 Just having his IP address is certainly not good enough (it's easy
 to
 spoof it), so we expect the client to sent some credentials in the
 first
 dedicated message. In LDAP this is done through a BindRequest.
 Anyway,
 as you require the user to send you some data, you have to process
 them
 by handling the messageReceived event.


 Last, not least : what about a separate thread ?

 That's a good question. The answer, again is it depends. If it
 takes
 seconds to authenticat a user, because you have to send the auth
 request
 to a remote server, then having a separate thread for that sounds
 smart.
 Most of the time, the authent will take a few ms, and will be done
 quite
 rarely, so it's enough to execute this code in the same thread.

 Hope it helps.

Re: User authentication

[Jon V.]

There really shouldn't be anything in created and opened unless you are
going to authenticate based on an IP address.

The authentication should be a part of your command message structure.

So you parse the commands into objects. Then you pass to the authentication
filter.

If message is of type authentication then do authentication else if
authenticated pass to the handler. If not authenticated and not type
authentication then close the session.

You only have two states. Authenticated or not.
On Aug 7, 2013 3:37 PM, Hunter McMillen mcmil...@gmail.com wrote:

 so all of the logic 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.

 In the message received just do...

 If ( authenticated )
 Next filter.message received( msg )

 Else

 Etc
 On Aug 7, 2013 3:22 PM, Hunter McMillen mcmil...@gmail.com wrote:

  So that is actually the solution we had before, but the state machine was
 in an authentication filter, we had a lot of trouble transitioning from
 ok
 now I am done authenticating, I want to be in my IoHandler now

 Hunter
 On 8/7/2013 3:03 PM, Jon wrote:

  So just implement some sort if authentication using the state machine
 and
 a database record lookup.

 Sent from my iPhone

 On Aug 7, 2013, at 2:11 PM, Hunter McMillen mcmil...@gmail.com wrote:

   I misunderstood what you were asking, yes we have an abstract Command

 class that has derivatives for all of the commands that a user could
 enter.

 Hunter
 On 8/7/2013 1:13 PM, Jon wrote:

  You must be building some sort of communication format for it.

 Sent from my iPhone

 On Aug 7, 2013, at 1:06 PM, Hunter McMillen mcmil...@gmail.com
 wrote:

   We're not really implementing a protocol, we are just trying to
 make a

 text-based game that will seamlessly support hundreds to thousands
 of users
 at a time.

 Hunter
 On 8/7/2013 12:02 PM, Jon wrote:

  This is more of a protocol implementation than a networking
 question.
   What kind of protocol are you implementing.

 Sent from my iPhone

 On Aug 7, 2013, at 11:34 AM, Hunter McMillen mcmil...@gmail.com
 wrote:

   We actually won't need anything as complex as LDAP, just a simple
 DB

 hash + salt lookup.

 Hunter
 On 8/7/2013 10:26 AM, Emmanuel Lécharny wrote:

  Le 8/7/13 4:11 PM, Hunter McMillen a écrit :

  What are the common ways to authenticate users using Mina?

 The first attempt I made was using a state machine, but I had
 problems
 integrating that with an IoHandler.

 In retrospect, the state machine seems like overkill; so I was
 hoping
 to get ideas for other ways to authenticate users, or maybe a
 link
 to
 an application that does some authentication.

 would it be terrible to do something like this?

 public void sessionCreated(IoSession session) {
   authenticateUser()  //  - Good idea? Bad Idea? Run in
 a
 separate thread?
 }

  Depends...

 let's see how it works with LDAP :

 - the user can connect on the server (and the sessionCreated event
 is
 handled), but the user will not be authentified at this point.

 - in order to authenticate the user, we need to know about the
 user.
 Just having his IP address is certainly not good enough (it's easy
 to
 spoof it), so we expect the client to sent some credentials in the
 first
 dedicated message. In LDAP this is done through a BindRequest.
 Anyway,
 as you require the user to send you some data, you have to process
 them
 by handling the messageReceived event.


 Last, not least : what about a separate thread ?

 That's a good question. The answer, again is it depends. If it
 takes
 seconds to authenticat a user, because you have to send the auth
 request
 to a remote server, then having a separate thread for that sounds
 smart.
 Most of the time, the authent will take a few ms, and will be done
 quite
 rarely, so it's enough to execute this code in the same thread.

 Hope it helps.

Re: Using an Authentication Filter and IoHandlerAdapter

[Jon V.]

Write your state to the IoSession and read it from the handler.
On Aug 1, 2013 10:06 PM, Hunter McMillen mcmil...@gmail.com wrote:

 Ok I think I understand the ordering now, we probably don't need an
 Executor. Our main issue is still how to link the state machine with our
 IoHandlerAdaptor.

 As of right now, after a user is authenticated (by the statemachine),
 the sessionIdle() event of the statemachine calls sessionOpened of our
 IoHandlerAdapter; obviously this is a logic error because sessionIdle()
 will fire frequently.

 So two things:
 1) Is there a way to leave the statemachine after we authenticate someone?
 2) If not, how to map the finishing of authentication to our
 IoHandlerAdapter ,specifically we probably don't want to fire a
 sessionCreated() event (because a session was already created by the
 statemachine), we simply want to pass that to our handler.

 Thanks.
 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, with Executor filter ordering is not guaranteed. Do you
  need an executor filter there?
  It's usage is recommended if you have some intensive task in chain or in
  IoHandler.
 
 
  Io-prorocol-authentication-executor-handler
 
  This sound good. Alternative is to use OrderedThreadPoolExecutor, but go
  simple first.
  Implement stuff without Executor filter and then push that in if needed.
 
 
  You cannot lock on out of order messaging without a queue and attempt to
  re-order the messages.
  On Aug 1, 2013 9:33 PM, Hunter McMillen mcmil...@gmail.com wrote:
 
  Sorry, the code is probably more useful to see, here is the entry point
  to our application:
 
  acceptor = new NioSocketAcceptor(
  Runtime.getRuntime().availableProcessors() );
  acceptor.getFilterChain().addLast(executor, new
  ExecutorFilter( Executors.newCachedThreadPool()));
  acceptor.getFilterChain().addLast(logger,
  MudConfig.Logging.getFilter());
  acceptor.getFilterChain().addLast(codec,
  new ProtocolCodecFilter(
  new TextLineEncoder(), new CommandDecoder()
  )
  );
 
  More importantly, my main question is how I can link the DONE state of
  the state machine (AuthenticationHandler) and the IoHandlerAdapter. I
  can post the code for these also, I just didn't want to overload the
  thread.
 
  Thanks.
  Hunter
  On 8/1/13 5:47 PM, Jon wrote:
  You are not using an executor filter right? You have to implement
  locking during the authentication phase if you are using thread
  scheduling.
  Sent from my iPhone
 
  On Aug 1, 2013, at 5:31 PM, Hunter McMillen mcmil...@gmail.com
  wrote:
  Hello,
 
  I recently started working on a project with a friend that is a
  text-based game. We were having trouble with ReadFuture's when trying
 to
  get a username/password combination from the user so we decided to
 follow
  the state machine example from Tapedeck TCP server on Mina's homepage:
 
 
 http://mina.apache.org/mina-project/xref/org/apache/mina/example/tapedeck/
  I have gotten the state machine to a point where it seems to be
  working
  well. It starts, reads a username, then a password, then has some logic
  to
  restart based on error; or it prints a message 'Authenticated'.
  However our main application logic is going to be (our plan at least)
  held in an IoHandlerAdapter, my question is what is a good way to
  integrate
  the two of these:
  1) The state machine authentication filter from the example above
  2) An IoHandlerAdapter that will track information about connected
  users and sessions
  My confusion mainly lies in how to transition between the state
  machine
  and the IoHandlerAdapter since they both respond to /sessionCreated
 /and
  /sessionOpened /events.
  Any help, ideas, or input would be appreciated.
 
  Thanks.
  Hunter
 
 
 

Re: Using an Authentication Filter and IoHandlerAdapter

[Jon V.]

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 not guarantee any kind of order.
  This means that you should implement the authentication phase before the
  executor.
 

 This is correct, with Executor filter ordering is not guaranteed. Do you
 need an executor filter there?
 It's usage is recommended if you have some intensive task in chain or in
 IoHandler.


 
  Io-prorocol-authentication-executor-handler
 

 This sound good. Alternative is to use OrderedThreadPoolExecutor, but go
 simple first.
 Implement stuff without Executor filter and then push that in if needed.


 
  You cannot lock on out of order messaging without a queue and attempt to
  re-order the messages.
  On Aug 1, 2013 9:33 PM, Hunter McMillen mcmil...@gmail.com wrote:
 
   Sorry, the code is probably more useful to see, here is the entry point
   to our application:
  
   acceptor = new NioSocketAcceptor(
   Runtime.getRuntime().availableProcessors() );
   acceptor.getFilterChain().addLast(executor, new
   ExecutorFilter( Executors.newCachedThreadPool()));
   acceptor.getFilterChain().addLast(logger,
   MudConfig.Logging.getFilter());
   acceptor.getFilterChain().addLast(codec,
   new ProtocolCodecFilter(
   new TextLineEncoder(), new CommandDecoder()
   )
   );
  
   More importantly, my main question is how I can link the DONE state of
   the state machine (AuthenticationHandler) and the IoHandlerAdapter. I
   can post the code for these also, I just didn't want to overload the
   thread.
  
   Thanks.
   Hunter
   On 8/1/13 5:47 PM, Jon wrote:
You are not using an executor filter right? You have to implement
   locking during the authentication phase if you are using thread
  scheduling.
   
Sent from my iPhone
   
On Aug 1, 2013, at 5:31 PM, Hunter McMillen mcmil...@gmail.com
  wrote:
   
Hello,
   
I recently started working on a project with a friend that is a
   text-based game. We were having trouble with ReadFuture's when trying
 to
   get a username/password combination from the user so we decided to
 follow
   the state machine example from Tapedeck TCP server on Mina's homepage:
  
 
 http://mina.apache.org/mina-project/xref/org/apache/mina/example/tapedeck/
   
I have gotten the state machine to a point where it seems to be
  working
   well. It starts, reads a username, then a password, then has some logic
  to
   restart based on error; or it prints a message 'Authenticated'.
   
However our main application logic is going to be (our plan at
 least)
   held in an IoHandlerAdapter, my question is what is a good way to
  integrate
   the two of these:
   
1) The state machine authentication filter from the example above
2) An IoHandlerAdapter that will track information about connected
   users and sessions
   
My confusion mainly lies in how to transition between the state
  machine
   and the IoHandlerAdapter since they both respond to /sessionCreated
 /and
   /sessionOpened /events.
   
Any help, ideas, or input would be appreciated.
   
Thanks.
Hunter
  
  
 



 --
 thanks
 ashish

 Blog: http://www.ashishpaliwal.com/blog
 My Photo Galleries: http://www.pbase.com/ashishpaliwal

Re: Using an Authentication Filter and IoHandlerAdapter

[Jon V.]

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.
On Aug 1, 2013 9:33 PM, Hunter McMillen mcmil...@gmail.com wrote:

 Sorry, the code is probably more useful to see, here is the entry point
 to our application:

 acceptor = new NioSocketAcceptor(
 Runtime.getRuntime().availableProcessors() );
 acceptor.getFilterChain().addLast(executor, new
 ExecutorFilter( Executors.newCachedThreadPool()));
 acceptor.getFilterChain().addLast(logger,
 MudConfig.Logging.getFilter());
 acceptor.getFilterChain().addLast(codec,
 new ProtocolCodecFilter(
 new TextLineEncoder(), new CommandDecoder()
 )
 );

 More importantly, my main question is how I can link the DONE state of
 the state machine (AuthenticationHandler) and the IoHandlerAdapter. I
 can post the code for these also, I just didn't want to overload the
 thread.

 Thanks.
 Hunter
 On 8/1/13 5:47 PM, Jon wrote:
  You are not using an executor filter right? You have to implement
 locking during the authentication phase if you are using thread scheduling.
 
  Sent from my iPhone
 
  On Aug 1, 2013, at 5:31 PM, Hunter McMillen mcmil...@gmail.com wrote:
 
  Hello,
 
  I recently started working on a project with a friend that is a
 text-based game. We were having trouble with ReadFuture's when trying to
 get a username/password combination from the user so we decided to follow
 the state machine example from Tapedeck TCP server on Mina's homepage:
 http://mina.apache.org/mina-project/xref/org/apache/mina/example/tapedeck/
 
  I have gotten the state machine to a point where it seems to be working
 well. It starts, reads a username, then a password, then has some logic to
 restart based on error; or it prints a message 'Authenticated'.
 
  However our main application logic is going to be (our plan at least)
 held in an IoHandlerAdapter, my question is what is a good way to integrate
 the two of these:
 
  1) The state machine authentication filter from the example above
  2) An IoHandlerAdapter that will track information about connected
 users and sessions
 
  My confusion mainly lies in how to transition between the state machine
 and the IoHandlerAdapter since they both respond to /sessionCreated /and
 /sessionOpened /events.
 
  Any help, ideas, or input would be appreciated.
 
  Thanks.
  Hunter

RE: Does MINA 2.0.4 supports IPV6

[Jon V.]

I am not setup to do anything in Mina at the moment.

You know how to setup regex patterns in Java?

Its as simple as converting the inetaddress to a string and comparing with
the regex.

If condition is true, close the session.
On Jul 30, 2013 1:17 AM, Nitin Phuria nit...@integramicro.com wrote:

 No I am not using any wildcards in blacklist filter of MINA. All the IP
 Addresses are IPV4 type and are individual IPs. Could you pls give me some
 leads on your statement You could build your own string and pattern based
 blacklist filter for IPV6 so that I can try it.

 Thanks And Regards,
 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
 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.
  https://issues.apache.org/jira/browse/DIRMINA-957
 
  Thanks And Regards,
  Nitin Phuria
 
  -Original Message-
  From: Emmanuel Lécharny [mailto:elecha...@gmail.com]
  Sent: Thursday, July 25, 2013 6:49 PM
  To: users@mina.apache.org
  Subject: Re: Does MINA 2.0.4 supports IPV6
 
  Le 7/25/13 1:47 PM, Nitin Phuria a écrit :
   Hello All,
  
   MINA build in BlacklistFilter does not support IPV6 address. If MINA
  itself
   supports binding to IPV4 and IPV6 why this restriction was kept in
   BlacklistFilter.
 
  That's a different problem...
 
 
   I checked the org.apache.mina.filter.firewall.Subnet class
   in MINA sources and found that it specifically checks for Inet4Address
  
   if (!(subnet instanceof Inet4Address))
   {
throw new IllegalArgumentException(Only IPv4 supported);
   }
  
   We are using this filter and now supports only IPV4
  
   Pls advice.
 
  Please fill a JIRA, we are going to look at it.
 
  ATM, I suggest you remove this check from the code, and see if it works
  for you. If so, we could then patch MINA, this is probably the fastest
  way to solve your issue.
 
 
  --
  Regards,
  Cordialement,
  Emmanuel Lécharny
  www.iktek.com
 
 
 

Re: Does MINA 2.0.4 supports IPV6

[Jon V.]

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.
 https://issues.apache.org/jira/browse/DIRMINA-957

 Thanks And Regards,
 Nitin Phuria

 -Original Message-
 From: Emmanuel Lécharny [mailto:elecha...@gmail.com]
 Sent: Thursday, July 25, 2013 6:49 PM
 To: users@mina.apache.org
 Subject: Re: Does MINA 2.0.4 supports IPV6

 Le 7/25/13 1:47 PM, Nitin Phuria a écrit :
  Hello All,
 
  MINA build in BlacklistFilter does not support IPV6 address. If MINA
 itself
  supports binding to IPV4 and IPV6 why this restriction was kept in
  BlacklistFilter.

 That's a different problem...


  I checked the org.apache.mina.filter.firewall.Subnet class
  in MINA sources and found that it specifically checks for Inet4Address
 
  if (!(subnet instanceof Inet4Address))
  {
   throw new IllegalArgumentException(Only IPv4 supported);
  }
 
  We are using this filter and now supports only IPV4
 
  Pls advice.

 Please fill a JIRA, we are going to look at it.

 ATM, I suggest you remove this check from the code, and see if it works
 for you. If so, we could then patch MINA, this is probably the fastest
 way to solve your issue.


 --
 Regards,
 Cordialement,
 Emmanuel Lécharny
 www.iktek.com

Rules

[Jon V.]

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

Re: XML Protocol with initial ASCII handshake

[Jon V.]

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
max.lars...@facilityboss.bizwrote:

 Hi Emmanuel,

 i have now a IoFilter which implement the initial connection
 establishing. But i have two question.

 1) In my filter how do i send response to a messageRecived, so that
 the IoHandler does get informed.

 2) The responses are only flushed out, during application shutdown.
 (J used flip() before sending the buffer via a DefaultWriteRequest
 down the nextFilter with filterWrite.

 I think the two questions relates somehow?

 regards

 Max


 On 18.04.13 02:00, Emmanuel Lécharny wrote:

 Le 4/17/13 7:56 PM, Max Larsson a écrit :

 Hi Emmanuel,

 Somehow i habe completly missed The concept of of filters. You are right
 They seem to fit perfectly. Just a question are iohanders even iofilters?

 No. They are two different interfaces. The IoHandler just process the
 events that are being generated by the IoService and propagated through
 the IoFilters.




 --
 __**__**_

 Max Larsson
 facilityboss GmbH
 Rheinstrasse 75, 64295 Darmstadt / Germany
 Handelsregister Darmstadt, HRB 86193
 Geschäftsführer: Dipl.-Inform. Max Lars Robert Larsson

 Mobil: +49 179 2184428
 Email: max.lars...@facilityboss.biz

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

[Jon V]

+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 side 
  byte
  Array is changed like [56, 34, 0, 4, -17, -65, -67, 1]. but when i transmit 
  byte
  array like [56,34,0,4,78,0] from one device to another using same approach 
  it
  will receive same byte array.
   
  Can you suggest me how to resolve the issue?
 No.
  
 Your desciption is at best extremely vague. It's like if you are asking
 why a plane has crashed without giving us nothing but a picture of the
 crashed plane...
  
  
  
  
 --  
 Regards,
 Cordialement,
 Emmanuel Lécharny
 www.iktek.com (http://www.iktek.com)  
  
  

Fw: Session Closed

[Jon V]

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,  

   I am trying to find out how Mina is detecting Session closures for 2.0.7. 
   Source searches do not seem to help. Can you point me in the right 
   direction?  
   
  Depends.
   
  - If the session is cleanly closed, your IoHandler will received a
  SessionClosed event. You just have to implement the sessionClosed(
  IoSession ) method in your IoHandler.
   
  - If the session is brutally closed, like when you unplug your network
  cable from the socket, the remote peer will never be informed. the only
  solutio is to check for the session idleness. When a session becomes
  idle (the idle delay is configurable), you will received a SessionIdle
  event. You have to implement the sessionIdle( IoSession ) method to trap
  this event and decide if the session has to be deleted.
   
  Is this what you are looking for ?
   
   
  --  
  Regards,
  Cordialement,
  Emmanuel Lécharny
  www.iktek.com (http://www.iktek.com)  
   
   
   
  
  

Mina Core 2.0.7 build fails from source zip

[Jon V.]

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]

[INFO] Total time: 1:36.140s
[INFO] Finished at: Thu Jan 03 10:44:28 EST 2013
[INFO] Final Memory: 18M/486M
[INFO]

[ERROR] Failed to execute goal
org.apache.maven.plugins:maven-surefire-plugin:2.12.2:test (default-test)
on project mina-core: There are test failures.
[ERROR]
[ERROR] Please refer to
C:\Users\Administrator\Downloads\apache-mina-2.0.7-src
(2)\apache-mina-2.0.7\src\mina-core\target\surefire-reports for the
individual test results.
[ERROR] - [Help 1]
[ERROR]
[ERROR] To see the full stack trace of the errors, re-run Maven with the -e
switch.
[ERROR] Re-run Maven using the -X switch to enable full debug logging.
[ERROR]
[ERROR] For more information about the errors and possible solutions,
please read the following articles:
[ERROR] [Help 1]
http://cwiki.apache.org/confluence/display/MAVEN/MojoFailureException
C:\Users\Administrator\Downloads\apache-mina-2.0.7-src
(2)\apache-mina-2.0.7\src\mina-core

Running org.apache.mina.transport.socket.nio.SocketConnectorTest
8,1,org.apache.mina.transport.socket.nio.SocketConnectorTest,testConnectFutureFailureTiming(org.apache.mina.transport.socket.nio.SocketConnectorTest),null,null,null,null,java.lang.AssertionError\r\n\tat
org.junit.Assert.fail(Assert.java:92)\r\n\tat
org.junit.Assert.fail(Assert.java:100)\r\n\tat
org.apache.mina.transport.AbstractConnectorTest.testConnectFutureFailureTiming(AbstractConnectorTest.java:121)\r\n\tat
sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)\r\n\tat
sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)\r\n\tat
sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)\r\n\tat
java.lang.reflect.Method.invoke(Method.java:601)\r\n\tat
org.junit.runners.model.FrameworkMethod$1.runReflectiveCall(FrameworkMethod.java:45)\r\n\tat
org.junit.internal.runners.model.ReflectiveCallable.run(ReflectiveCallable.java:15)\r\n\tat
org.junit.runners.model.FrameworkMethod.invokeExplosively(FrameworkMethod.java:42)\r\n\tat
org.junit.internal.runners.statements.InvokeMethod.evaluate(InvokeMethod.java:20)\r\n\tat
org.junit.runners.ParentRunner.runLeaf(ParentRunner.java:263)\r\n\tat
org.junit.runners.BlockJUnit4ClassRunner.runChild(BlockJUnit4ClassRunner.java:68)\r\n\tat
org.junit.runners.BlockJUnit4ClassRunner.runChild(BlockJUnit4ClassRunner.java:47)\r\n\tat
org.junit.runners.ParentRunner$3.run(ParentRunner.java:231)\r\n\tat
org.junit.runners.ParentRunner$1.schedule(ParentRunner.java:60)\r\n\tat
org.junit.runners.ParentRunner.runChildren(ParentRunner.java:229)\r\n\tat
org.junit.runners.ParentRunner.access$000(ParentRunner.java:50)\r\n\tat
org.junit.runners.ParentRunner$2.evaluate(ParentRunner.java:222)\r\n\tat
org.junit.runners.ParentRunner.run(ParentRunner.java:300)\r\n\tat
org.apache.maven.surefire.junit4.JUnit4Provider.execute(JUnit4Provider.java:252)\r\n\tat
org.apache.maven.surefire.junit4.JUnit4Provider.executeTestSet(JUnit4Provider.java:141)\r\n\tat
org.apache.maven.surefire.junit4.JUnit4Provider.invoke(JUnit4Provider.java:112)\r\n\tat
sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)\r\n\tat
sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)\r\n\tat
sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)\r\n\tat
java.lang.reflect.Method.invoke(Method.java:601)\r\n\tat
org.apache.maven.surefire.util.ReflectionUtils.invokeMethodWithArray(ReflectionUtils.java:189)\r\n\tat
org.apache.maven.surefire.booter.ProviderFactory$ProviderProxy.invoke(ProviderFactory.java:165)\r\n\tat
org.apache.maven.surefire.booter.ProviderFactory.invokeProvider(ProviderFactory.java:85)\r\n\tat
org.apache.maven.surefire.booter.ForkedBooter.runSuitesInProcess(ForkedBooter.java:113)\r\n\tat
org.apache.maven.surefire.booter.ForkedBooter.main(ForkedBooter.java:75)\r\n
5,1,org.apache.mina.transport.socket.nio.SocketConnectorTest,testConnectFutureSuccessTiming(org.apache.mina.transport.socket.nio.SocketConnectorTest),null,null,null
6,1,org.apache.mina.transport.socket.nio.SocketConnectorTest,testConnectFutureSuccessTiming(org.apache.mina.transport.socket.nio.SocketConnectorTest),null,null,null
5,1,org.apache.mina.transport.socket.nio.SocketConnectorTest,testSessionCallbackInvocation(org.apache.mina.transport.socket.nio.SocketConnectorTest),null,null,null
6,1,org.apache.mina.transport.socket.nio.SocketConnectorTest,testSessionCallbackInvocation(org.apache.mina.transport.socket.nio.SocketConnectorTest),null,null,null
2,1,org.apache.maven.surefire.junit4.JUnit4Provider,org.apache.mina.transport.socket.nio.SocketConnectorTest,null,null,null

Patch for IoBuffer management in IoProcessor 2.0.7

[Jon V.]

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 implementation that I will be happy to
submit back once I have tested it properly.

Re: Patch for IoBuffer management in IoProcessor 2.0.7

[Jon V.]

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() to release the
 IoBuffer
  memory.
 
  The CachedBufferAllocator is totally broken but I am able to get 100%
  recycled Buffers with a custom implementation that I will be happy to
  submit back once I have tested it properly.
 
 Hi !

 the patch has been discarded by the mailer.

 Could you create a JIRA and attach the patch ?

 Many thanks !

 --
 Regards,
 Cordialement,
 Emmanuel Lécharny
 www.iktek.com

Re: Patch for IoBuffer management in IoProcessor 2.0.7

[Jon V.]

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...@gmail.com
 wrote:
 
  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() to release the
  IoBuffer
   memory.
  
   The CachedBufferAllocator is totally broken but I am able to get 100%
   recycled Buffers with a custom implementation that I will be happy to
   submit back once I have tested it properly.
  
  Hi !
 
  the patch has been discarded by the mailer.
 
  Could you create a JIRA and attach the patch ?
 
  Many thanks !
 
  --
  Regards,
  Cordialement,
  Emmanuel Lécharny
  www.iktek.com
 
 

Re: FilterChain instantiation

[Jon V]

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 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 one? (next polling interval)
 
 both the decoder and encoder must be monotomic.  I need to know if I should 
 plan for enforcing single thread processing. 

Re: FilterChain instantiation

[Jon V]

Emanuel, 

Thank you for the reply.

Since the mailing list system was not sending me replies, I was able to gain 
access to the replies through the apache mail archives system.

I wrote monotomic which is a mis-spelling (lets blame it on the 
auto-correction in Mac OSX) for monotonic.  Monotonic, by nature means that 
an ordering mechanism must always be progressing forward and in 
order.(guaranteed order) e.g. 1,2,3,4,5

I will expand on my second question.

1. Lets say that there are 4 polling threads and one session.

2. Mina fires off messageReceived #1 during one polling cycle.  Protocol chain 
is executing.

3. Another polling cycle comes around, the session has new information in the 
buffer.

4. Will Mina fire off the new buffer while the previous cycle is still 
executing?  say it takes longer than the polling cycle to complete processing 
of the data.  Since my protocol 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.  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 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 one? (next polling interval)
  
  both the decoder and encoder must be monotomic.  I need to know if I should 
  plan for enforcing single thread processing. 
 

FilterChain instantiation

[Jon V]

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 one? (next polling interval)

both the decoder and encoder must be monotomic.  I need to know if I should 
plan for enforcing single thread processing.