Nabble - Netty Developer Group

[JBoss JIRA] Created: (NETTY-104) Rename org.jboss.netty.group to org.jboss.netty.channel.group

2009-01-07T21:48:04Z

Rename org.jboss.netty.group to org.jboss.netty.channel.group
-------------------------------------------------------------

Key: NETTY-104
URL: https://jira.jboss.org/jira/browse/NETTY-104
Project: Netty
Issue Type: Task
Components: Core API
Reporter: Trustin Lee
Assignee: Trustin Lee
Priority: Trivial
Fix For: 3.1.0.ALPHA3

The group API is for Channels, so it should reside under the org.jboss.netty.channel package.

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

[JBoss JIRA] Closed: (NETTY-104) Rename org.jboss.netty.group to org.jboss.netty.channel.group

2009-01-07T21:48:04Z

[ https://jira.jboss.org/jira/browse/NETTY-104?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Trustin Lee closed NETTY-104.
-----------------------------

Resolution: Done

Done.

> Rename org.jboss.netty.group to org.jboss.netty.channel.group
> -------------------------------------------------------------
>
> Key: NETTY-104
> URL: https://jira.jboss.org/jira/browse/NETTY-104
> Project: Netty
> Issue Type: Task
> Components: Core API
> Reporter: Trustin Lee
> Assignee: Trustin Lee
> Priority: Trivial
> Fix For: 3.1.0.ALPHA3
>
>
> The group API is for Channels, so it should reside under the org.jboss.netty.channel package.

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

Re: Google Protocol Buffers integration is ready.

2009-01-07T04:37:18Z

I find this page more convenient to browse the source code:

http://tinyurl.com/73nflb

— Trustin Lee, http://gleamynode.net/

On Wed, Jan 7, 2009 at 9:24 PM, Trustin Lee (이희승) <trustin@...> wrote:

Hi folks,

I have finished the implementation of Google Protocol Buffers codec for Netty.

Google Protocol Buffers (protobuf) are a way of encoding structured data in an efficient yet extensible format. Google uses Protocol Buffers for almost all of its internal RPC protocols and file formats. With protobuf, you can define a binary protocol very quickly and efficiently. For more information, please see here: http://code.google.com/p/protobuf/

With the ProtobufEncoder and ProtobufDecoder that Netty provides, you can now write a network application with a very efficient binary encoding even more quickly than ever. It just takes little time to implement a codec with protobuf.

To help your understanding, I have written a time client / server example. The time client sends a list of time zones, and then the time server responds with the local time for each time zone. Please brwose the following directory:

http://fisheye.jboss.org/browse/Netty/trunk/src/main/java/org/jboss/netty/example/localtime
Please note that LocalTimeProtocol.java has been generated from LocalTimeProtocol.proto by protobuf compiler.

Please feel free to send me a feed back once you review the source code. I'd like to know what could be improved to support protobuf better and to help you implement an efficient binary protocol more quickly than ever.

Thanks in advance,

— Trustin Lee, http://gleamynode.net/

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

Trustin Lee, Principal Software Engineer, JBoss, a division of Red Hat
--
what we call human nature is actually human habit
--
http://gleamynode.net/

Google Protocol Buffers integration is ready.

2009-01-07T04:24:53Z

Hi folks,

I have finished the implementation of Google Protocol Buffers codec for Netty.

Google Protocol Buffers (protobuf) are a way of encoding structured data in an efficient yet extensible format. Google uses Protocol Buffers for almost all of its internal RPC protocols and file formats. With protobuf, you can define a binary protocol very quickly and efficiently. For more information, please see here: http://code.google.com/p/protobuf/

With the ProtobufEncoder and ProtobufDecoder that Netty provides, you can now write a network application with a very efficient binary encoding even more quickly than ever. It just takes little time to implement a codec with protobuf.

To help your understanding, I have written a time client / server example. The time client sends a list of time zones, and then the time server responds with the local time for each time zone. Please brwose the following directory:

http://fisheye.jboss.org/browse/Netty/trunk/src/main/java/org/jboss/netty/example/localtime

Please note that LocalTimeProtocol.java has been generated from LocalTimeProtocol.proto by protobuf compiler.

Please feel free to send me a feed back once you review the source code. I'd like to know what could be improved to support protobuf better and to help you implement an efficient binary protocol more quickly than ever.

Thanks in advance,

— Trustin Lee, http://gleamynode.net/

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

Trustin Lee, Principal Software Engineer, JBoss, a division of Red Hat
--
what we call human nature is actually human habit
--
http://gleamynode.net/

[JBoss JIRA] Closed: (NETTY-103) Support the Android platform

2009-01-06T16:59:04Z

[ https://jira.jboss.org/jira/browse/NETTY-103?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Trustin Lee closed NETTY-103.
-----------------------------

Resolution: Done

Should run fine now - I used reflection to reference the classes in the java.lang.management package.

> Support the Android platform
> ----------------------------
>
> Key: NETTY-103
> URL: https://jira.jboss.org/jira/browse/NETTY-103
> Project: Netty
> Issue Type: Feature Request
> Components: Core API
> Reporter: Trustin Lee
> Assignee: Trustin Lee
> Priority: Minor
> Fix For: 3.1.0.ALPHA3
>
>
> It's fairly easy to support the Android platform according to my quick research. TimeBasedUuidGenerator is the only class that needs modification - it depends on javax.management which doesn't exist in Android.

[JBoss JIRA] Created: (NETTY-103) Support the Android platform

2009-01-06T16:37:04Z

Support the Android platform
----------------------------

Key: NETTY-103
URL: https://jira.jboss.org/jira/browse/NETTY-103
Project: Netty
Issue Type: Feature Request
Components: Core API
Reporter: Trustin Lee
Assignee: Trustin Lee
Priority: Minor
Fix For: 3.1.0.ALPHA3

It's fairly easy to support the Android platform according to my quick research. TimeBasedUuidGenerator is the only class that needs modification - it depends on javax.management which doesn't exist in Android.

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

[JBoss JIRA] Closed: (NETTY-102) Allow DelimiterBasedFrameDecoder to include the delimiter in the frame.

2009-01-05T21:13:05Z

[ https://jira.jboss.org/jira/browse/NETTY-102?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Trustin Lee closed NETTY-102.
-----------------------------

Resolution: Done

Done. Now you can specify whether to strip the delimiter from the decoded frame or not in the constructor.

> Allow DelimiterBasedFrameDecoder to include the delimiter in the frame.
> -----------------------------------------------------------------------
>
> Key: NETTY-102
> URL: https://jira.jboss.org/jira/browse/NETTY-102
> Project: Netty
> Issue Type: Feature Request
> Components: Handler
> Reporter: Trustin Lee
> Assignee: Trustin Lee
> Priority: Minor
> Fix For: 3.1.0.ALPHA3
>
>
> It's often useful to figure out what delimiter generated a frame. For example, a server might want to use the same line delimiter style (DOS or UNIX) with the line delimiter style of a client.

[JBoss JIRA] Created: (NETTY-102) Allow DelimiterBasedFrameDecoder to include the delimiter in the frame.

2009-01-05T20:43:14Z

Allow DelimiterBasedFrameDecoder to include the delimiter in the frame.
-----------------------------------------------------------------------

Key: NETTY-102
URL: https://jira.jboss.org/jira/browse/NETTY-102
Project: Netty
Issue Type: Feature Request
Components: Handler
Reporter: Trustin Lee
Assignee: Trustin Lee
Priority: Minor
Fix For: 3.1.0.ALPHA3

It's often useful to figure out what delimiter generated a frame. For example, a server might want to use the same line delimiter style (DOS or UNIX) with the line delimiter style of a client.

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

[JBoss JIRA] Created: (NETTY-101) ObjectDecoderInputStream.readObject() discards data and throws a StreamCorruptedException.

2009-01-05T18:07:04Z

ObjectDecoderInputStream.readObject() discards data and throws a StreamCorruptedException.
------------------------------------------------------------------------------------------

Key: NETTY-101
URL: https://jira.jboss.org/jira/browse/NETTY-101
Project: Netty
Issue Type: Bug
Components: Handler
Reporter: Trustin Lee
Assignee: Trustin Lee
Fix For: 3.1.0.ALPHA3

ObjectDecoderInputStream.readObject() looks like this:

int dataLen = readInt();
byte[] data = new byte[dataLen];
readFully(data);
return new CompactObjectInputStream(in, classLoader).readObject();

and it is obviously wrong because the array 'data' is not used at all after consuming the serialized data.

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

[JBoss JIRA] Closed: (NETTY-101) ObjectDecoderInputStream.readObject() discards data and throws a StreamCorruptedException.

2009-01-05T18:07:04Z

[ https://jira.jboss.org/jira/browse/NETTY-101?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Trustin Lee closed NETTY-101.
-----------------------------

Resolution: Done

Fixed.

> ObjectDecoderInputStream.readObject() discards data and throws a StreamCorruptedException.
> ------------------------------------------------------------------------------------------
>
> Key: NETTY-101
> URL: https://jira.jboss.org/jira/browse/NETTY-101
> Project: Netty
> Issue Type: Bug
> Components: Handler
> Reporter: Trustin Lee
> Assignee: Trustin Lee
> Fix For: 3.1.0.ALPHA3
>
>
> ObjectDecoderInputStream.readObject() looks like this:
> int dataLen = readInt();
> byte[] data = new byte[dataLen];
> readFully(data);
> return new CompactObjectInputStream(in, classLoader).readObject();
> and it is obviously wrong because the array 'data' is not used at all after consuming the serialized data.

[JBoss JIRA] Created: (NETTY-100) Add ExecutionHandler.terminateExecutor()

2009-01-02T08:48:54Z

Add ExecutionHandler.terminateExecutor()
----------------------------------------

Key: NETTY-100
URL: https://jira.jboss.org/jira/browse/NETTY-100
Project: Netty
Issue Type: Feature Request
Components: Handler
Reporter: Trustin Lee
Assignee: Trustin Lee
Priority: Optional
Fix For: 3.1.0.ALPHA3

It would be more convenient if a user can shut down and terminate the ExecutorService conveniently which was specified as a constructor parmeter of ExecutionHandler, just like we do with ChannelFactory.releaseExternalResources()

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

[JBoss JIRA] Closed: (NETTY-100) Add ExecutionHandler.terminateExecutor()

2009-01-02T08:48:54Z

[ https://jira.jboss.org/jira/browse/NETTY-100?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Trustin Lee closed NETTY-100.
-----------------------------

Resolution: Done

Done.

> Add ExecutionHandler.terminateExecutor()
> ----------------------------------------
>
> Key: NETTY-100
> URL: https://jira.jboss.org/jira/browse/NETTY-100
> Project: Netty
> Issue Type: Feature Request
> Components: Handler
> Reporter: Trustin Lee
> Assignee: Trustin Lee
> Priority: Optional
> Fix For: 3.1.0.ALPHA3
>
>
> It would be more convenient if a user can shut down and terminate the ExecutorService conveniently which was specified as a constructor parmeter of ExecutionHandler, just like we do with ChannelFactory.releaseExternalResources()

[JBoss JIRA] Commented: (NETTY-99) Add the links to the projects which depend on Netty

2008-12-31T21:58:54Z

[ https://jira.jboss.org/jira/browse/NETTY-99?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12444355#action_12444355 ]

Richard Jackson commented on NETTY-99:
--------------------------------------

Just starting a project that will use Netty for both client and server tasks.

Project scope right now: Implement the OSGi HTTPService specification

Project home : http://code.google.com/p/osgimodules/

Not much there yet as I'm just starting on this.

> Add the links to the projects which depend on Netty
> ---------------------------------------------------
>
> Key: NETTY-99
> URL: https://jira.jboss.org/jira/browse/NETTY-99
> Project: Netty
> Issue Type: Task
> Reporter: Trustin Lee
> Assignee: Trustin Lee
> Priority: Optional
>
> It would be nice if there's a dedicated page that contains:
> * a link to a project that depends on Netty
> * a link to a specific source code file / directory of the project that uses Netty

[JBoss JIRA] Created: (NETTY-99) Add the links to the projects which depend on Netty

2008-12-30T21:09:54Z

Add the links to the projects which depend on Netty
---------------------------------------------------

Key: NETTY-99
URL: https://jira.jboss.org/jira/browse/NETTY-99
Project: Netty
Issue Type: Task
Reporter: Trustin Lee
Assignee: Trustin Lee
Priority: Optional

It would be nice if there's a dedicated page that contains:

* a link to a project that depends on Netty
* a link to a specific source code file / directory of the project that uses Netty

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

[JBoss JIRA] Created: (NETTY-98) Annotation-based state machine which helps ChannelHandler implementation

2008-12-29T04:06:55Z

Annotation-based state machine which helps ChannelHandler implementation
------------------------------------------------------------------------

Key: NETTY-98
URL: https://jira.jboss.org/jira/browse/NETTY-98
Project: Netty
Issue Type: Feature Request
Components: Handler
Reporter: Trustin Lee
Assignee: Trustin Lee
Fix For: 3.2.0.ALPHA1

Long awaited feature that should be implemented:

http://n2.nabble.com/New-Feature-Request---State-Machine-td1563216.html

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

Updated road map for 3.1 (and 3.2)

2008-12-29T04:02:13Z

Hi folks,

I've just updated the road map for Netty 3.1. Please note that some
issues with lower priority were rescheduled to 3.2:

* https://jira.jboss.org/jira/secure/BrowseProject.jspa?id=12310721&subset=-1

As you see, remaining noteworthy to-dos for 3.1 are:

* IoC container integration (3.1.ALPHA3) - JBoss Microcontainer,
Google Guice, and Spring
* Google Protocol Buffers integration (3.1.ALPHA3)
* Idleness notification (3.1.ALPHA4)
* XNIO transport (3.1.ALPHA4)

ALPHA3 has been delayed quite a lot, but I hope it will be released
before Jan 4, 2009. Please stay tuned!

At last but not least, please feel free to propose a feature which
would be nice to have - it's not too late. :-)

Cheers,
Trustin
--
Trustin Lee, Principal Software Engineer, JBoss, a division of Red Hat
--
what we call human nature in actuality is human habit
--
http://gleamynode.net/
_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

Trustin Lee, Principal Software Engineer, JBoss, a division of Red Hat
--
what we call human nature is actually human habit
--
http://gleamynode.net/

Re: Integration with Google Protocol Buffer

2008-12-17T04:21:57Z

Asked the same question to the protobuf official forum:

* http://tinyurl.com/6faayl

Feel free to jump in!

On Wed, Dec 17, 2008 at 09:12:07PM +0900, Trustin Lee wrote:

> Hi folks,
>
> A user one day told me that it would be nice if Netty has a codec for
> Google Protocol Buffers, and I think it's a great idea. Here's a JIRA
> issue:
>
> * https://jira.jboss.org/jira/browse/NETTY-97
>
> However, I don't have much experience with protobuf yet although I know
> how to encode and decode a protobuf message programmatically. From my
> understanding, it seems like a user usually defines a top level message
> type that can be a container of other messages. This approach is
> described here in detail:
>
> * Official documentation - http://tinyurl.com/56cerv
> * Forum discussion - http://tinyurl.com/5zabns
>
> Does anyone here have experience with protobuf in a fairly complex
> application? Any feed back would be appreciated!
>
> Cheers,
> Trustin
>

--
Trustin Lee, Principal Software Engineer, JBoss, a division of Red Hat
--
what we call human nature in actuality is human habit
--
http://gleamynode.net/

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

attachment0 (204 bytes) Download Attachment

Integration with Google Protocol Buffer

2008-12-17T04:09:55Z

Hi folks,

A user one day told me that it would be nice if Netty has a codec for
Google Protocol Buffers, and I think it's a great idea. Here's a JIRA
issue:

* https://jira.jboss.org/jira/browse/NETTY-97

However, I don't have much experience with protobuf yet although I know
how to encode and decode a protobuf message programmatically. From my
understanding, it seems like a user usually defines a top level message
type that can be a container of other messages. This approach is
described here in detail:

* Official documentation - http://tinyurl.com/56cerv
* Forum discussion - http://tinyurl.com/5zabns

Does anyone here have experience with protobuf in a fairly complex
application? Any feed back would be appreciated!

Cheers,
Trustin

--
Trustin Lee, Principal Software Engineer, JBoss, a division of Red Hat
--
what we call human nature in actuality is human habit
--
http://gleamynode.net/

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

attachment0 (204 bytes) Download Attachment

[JBoss JIRA] Closed: (NETTY-97) Google Protocol Buffer codec

2009-01-07T04:28:04Z

[ https://jira.jboss.org/jira/browse/NETTY-97?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Trustin Lee closed NETTY-97.
----------------------------

Resolution: Done

Done - http://n2.nabble.com/Google-Protocol-Buffers-integration-is-ready.-td2122078.html

> Google Protocol Buffer codec
> ----------------------------
>
> Key: NETTY-97
> URL: https://jira.jboss.org/jira/browse/NETTY-97
> Project: Netty
> Issue Type: Feature Request
> Components: Handler
> Reporter: Trustin Lee
> Assignee: Trustin Lee
> Fix For: 3.1.0.ALPHA3
>
>
> Out-of-the-box Google protobuf codec should enable rapid implementation of efficient binary message exchange. Moreover, it's very easy to implement. Another potential advantage is attraction of more users by adopting hot technology. :)
> [Slightly OT] I'm not sure yet, but Preon also looks interesting:
> * Preon - http://preon.sourceforge.net/

[JBoss JIRA] Commented: (NETTY-97) Google Protocol Buffer codec

2008-12-29T18:18:54Z

[ https://jira.jboss.org/jira/browse/NETTY-97?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12444236#action_12444236 ]

Trustin Lee commented on NETTY-97:
----------------------------------

Just finished the initial implementation - needs to write a working example.

[JBoss JIRA] Commented: (NETTY-97) Google Protocol Buffer codec

2008-12-17T04:09:54Z

[ https://jira.jboss.org/jira/browse/NETTY-97?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12443081#action_12443081 ]

Trustin Lee commented on NETTY-97:
----------------------------------

Related Q&A - http://tinyurl.com/6e9s42

[JBoss JIRA] Created: (NETTY-97) Google Protocol Buffer codec

2008-12-17T03:56:10Z

Google Protocol Buffer codec
----------------------------

Key: NETTY-97
URL: https://jira.jboss.org/jira/browse/NETTY-97
Project: Netty
Issue Type: Feature Request
Components: Handler
Reporter: Trustin Lee
Assignee: Trustin Lee
Fix For: 3.1.0.ALPHA3

Out-of-the-box Google protobuf codec should enable rapid implementation of efficient binary message exchange. Moreover, it's very easy to implement. Another potential advantage is attraction of more users by adopting hot technology. :)

[Slightly OT] I'm not sure yet, but Preon also looks interesting:

* Preon - http://preon.sourceforge.net/

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

Re: Blocking with channelGroup.close with awaitUninterruptibly

2008-12-17T03:51:05Z

Thanks for the quick update, too. :-)

Cheers,
Trustin

On Wed, Dec 17, 2008 at 08:29:21PM +0900, Frederic Bregier wrote:

>
> Hi Trustin,
>
> Thanks for the quick fix! It works!
> I've just updated the Wiki.
>
> Frederic
>
>
> -----
> Hardware/Software Architect

attachment0 (204 bytes) Download Attachment

Re: Blocking with channelGroup.close with awaitUninterruptibly

2008-12-17T03:29:21Z

Hi Trustin,

Thanks for the quick fix! It works!
I've just updated the Wiki.

Frederic

Hardware/Software Architect

Re: Blocking with channelGroup.close with awaitUninterruptibly

2008-12-17T01:39:12Z

Hi Frederic,

Thanks for reporting a bug. I was able to reproduce it from your
example and have just checked in the fix. The returned
ChannelGroupFuture didn't complete at all when there's no channel in the
group (because the connection is closed already.)

Now that this bug is fixed, I think the example should be modified to
call awaitUninterruptibly().

Cheers,
Trustin

On Wed, Dec 17, 2008 at 02:33:26AM +0900, Frederic Bregier wrote:

>
> Hi Trustin,
>
> I've got a short question (I guess).
> If you take the example from the wiki, you can easily reproduce the
> behaviour.
> It is true for the client and for the server.
> Here is the logic:
>
> In the main class:
> ChannelGroup channelGroup = ChabbelGroupFactory.getGroup(...);
> ClientBootstrap bootstrap = new ClientBootstrap(...);
> ... (adding the handler and the channelGroup to the handler)
> ChannelFuture future = bootstrap.connect(...).awaitUninterruptibly();
> Channel channel = future.getChannel();
> channelGroup.add(channel);
>
> In the Handler:
> void channelConnected(...) {
> channelGroup.add(ctx.getChannel();
> }
>
> Back into the client main class:
> channelGroup.close();
>
> In this case, no problem.
> But if I do:
> channelGroup.close().awaitUninterruptibly();
> It blocks forever...
>
> Do I missed something?
>
> Frederic
>
> -----
> Hardware/Software Architect

attachment0 (204 bytes) Download Attachment

Blocking with channelGroup.close with awaitUninterruptibly

2008-12-16T09:33:26Z

Hi Trustin,

I've got a short question (I guess).
If you take the example from the wiki, you can easily reproduce the behaviour.
It is true for the client and for the server.
Here is the logic:

In the main class:
ChannelGroup channelGroup = ChabbelGroupFactory.getGroup(...);
ClientBootstrap bootstrap = new ClientBootstrap(...);
... (adding the handler and the channelGroup to the handler)
ChannelFuture future = bootstrap.connect(...).awaitUninterruptibly();
Channel channel = future.getChannel();
channelGroup.add(channel);

In the Handler:
void channelConnected(...) {
channelGroup.add(ctx.getChannel();
}

Back into the client main class:
channelGroup.close();

In this case, no problem.
But if I do:
channelGroup.close().awaitUninterruptibly();
It blocks forever...

Do I missed something?

Frederic

Hardware/Software Architect

Wiki space has been created - please contribute your idea!

2008-12-15T21:02:13Z

Hi folks,

I've just created a wiki page in JBoss.org so that we can share user
contributed resources. You can leave comments on the existing resources
and add wiki pages to contribute to and interact with the community.

* http://www.jboss.org/community/docs/DOC-13205

I also added a link to the wiki page in our documentation section:
(See 'User contributed tutorials and examples.')

* http://www.jboss.org/netty/documentation.html

As the first stuff, I have added manish's Dieffie-Hellman Key Exchange
Server example. I think Frederic's object serialization example could
be added, too!

Alright, let's see how it grows up! :-)

Cheers,

--
Trustin Lee, Principal Software Engineer, JBoss, a division of Red Hat
--
what we call human nature in actuality is human habit
--
http://gleamynode.net/

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

attachment0 (204 bytes) Download Attachment

Re: exception during cleanup

2008-12-17T01:06:37Z

Hi,

Sorry for the late reply; somewhat distracted by various things.

On Fri, Dec 12, 2008 at 01:20:25AM +0900, jdv145 wrote:

> First some preliminary results of my usage of netty. I'm using it for a
> server to host a boardgave made with flash. It's working way better then
> expected and the only problems i have are pretty much only with flash. From
> a developer's perspective java is way way better then flash. This also means
> netty is working pretty much flawlessly. I will give a more elaborate
> report (with some problems i faced which might be handy for the noobs), but
> only after the game goes live (for now the game has pretty much only run in
> a test environment with 10 ppl at most and artificial load generation). You
> did a real good job. A lot of people which will be playing my game are
> thankful for it ;). Im glad i made the jump from mina to netty, mina is
> okay.. but i like netty better :).

Thank you very much for your ongoing feed back, and please keep us
updated so that we can improve Netty on and on!

> Anyway, i just got an exception. It doesn't happpen often and I cant seem to
> reproduce it (but i didn't really tried it either). Did i something wrong or
> is there a bug in netty? If you need a stackdump you have to help me a
> little cause i haven't made one ever :P.

ClosedChannelException is thrown when:

* you attempted to write a message to a closed channel
* you wrote a message to an open channel, but the channel was closed
before the message is written out

If you attempted to write a message to a closed channel, then your code
might have some timing issue. Otherwise, it's the case where the
connection was closed unexpectedly (e.g. remote peer closed the connection
without agreement.)

HTH,

--
Trustin Lee, Principal Software Engineer, JBoss, a division of Red Hat
--
what we call human nature in actuality is human habit
--
http://gleamynode.net/

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

attachment0 (204 bytes) Download Attachment

exception during cleanup

2008-12-11T08:20:25Z

First some preliminary results of my usage of netty. I'm using it for a server to host a boardgave made with flash. It's working way better then expected and the only problems i have are pretty much only with flash. From a developer's perspective java is way way better then flash. This also means netty is working pretty much flawlessly. I will give a more elaborate report (with some problems i faced which might be handy for the noobs), but only after the game goes live (for now the game has pretty much only run in a test environment with 10 ppl at most and artificial load generation). You did a real good job. A lot of people which will be playing my game are thankful for it ;). Im glad i made the jump from mina to netty, mina is okay.. but i like netty better :).

Anyway, i just got an exception. It doesn't happpen often and I cant seem to reproduce it (but i didn't really tried it either). Did i something wrong or is there a bug in netty? If you need a stackdump you have to help me a little cause i haven't made one ever :P.

11 Dec 2008 16:59:15,671 - ERROR New I/O server worker #1-1 gamehost.MessageDecoder - Invallid communication or other problem in the MessageDecoder. Ipnr: 127.0.0.1
java.nio.channels.ClosedChannelException
at org.jboss.netty.channel.socket.nio.NioWorker.cleanUpWriteBuffer(NioWorker.java:748)
at org.jboss.netty.channel.socket.nio.NioWorker.close(NioWorker.java:732)
at org.jboss.netty.channel.socket.nio.NioServerSocketPipelineSink.handleAcceptedSocket(NioServerSocketPipelineSink.java:121)
at org.jboss.netty.channel.socket.nio.NioServerSocketPipelineSink.eventSunk(NioServerSocketPipelineSink.java:78)
at org.jboss.netty.handler.codec.string.StringEncoder.handleDownstream(StringEncoder.java:93)
at org.jboss.netty.channel.Channels.close(Channels.java:736)
at org.jboss.netty.channel.AbstractChannel.close(AbstractChannel.java:158)
at org.jboss.netty.channel.ChannelFutureListener$1.operationComplete(ChannelFutureListener.java:45)
at org.jboss.netty.channel.DefaultChannelFuture.notifyListener(DefaultChannelFuture.java:311)
at org.jboss.netty.channel.DefaultChannelFuture.notifyListeners(DefaultChannelFuture.java:297)
at org.jboss.netty.channel.DefaultChannelFuture.setSuccess(DefaultChannelFuture.java:250)
at org.jboss.netty.channel.socket.nio.NioWorker.writeUnfair(NioWorker.java:460)
at org.jboss.netty.channel.socket.nio.NioWorker.write(NioWorker.java:411)
at org.jboss.netty.channel.socket.nio.NioSocketChannel$WriteTask.run(NioSocketChannel.java:127)
at org.jboss.netty.channel.socket.nio.NioWorker.processWriteTaskQueue(NioWorker.java:231)
at org.jboss.netty.channel.socket.nio.NioWorker.run(NioWorker.java:160)
at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
at java.lang.Thread.run(Thread.java:619)
11 Dec 2008 16:59:20,671 - INFO New I/O server worker #1-4 gamehost.Tools - Can't send to cha

Re: Exception under heavy load when doing Boss.register.

2008-12-12T02:57:20Z

Glad that it worked! Please keep us updated with your ongoing feed
back. :)

Cheers,
Trustin

On Fri, Dec 12, 2008 at 07:34:49AM +0900, Wesly smith wrote:

>
> Updated to the newest trunk (rev 608), punched with 500,000 connections, and
> the problem is gone.It is a nice fix, and thanks a lot for the work.
>
> On Thu, Dec 11, 2008 at 4:23 PM, wesly smith0 <weslysmith0@...> wrote:
>
>> I will follow up when I try the newest (rev 608 today) version on svn.
>>
>>
>> On Thu, Dec 11, 2008 at 12:40 AM, Trustin Lee-2 (via Nabble) <
>> ml-user+63047-178166325@...<ml-user%2B63047-178166325@...>
>>> wrote:
>>
>>> Could you try the latest revision in the Subversion repository? I have
>>> made significant improvement in NioClientSocketPipelineSink.
>>>
>>> HTH,
>>> Trustin
>>>
>>> On Thu, Dec 11, 2008 at 07:32:55AM +0900, Wesly smith wrote:
>>>>
>>>> Exception under heavy load when doing Boss.register, the stack trace is
>>> the
>>>> below.
>>>>
>>>> I saw those NPE under the load > 1500 connections/seconds for about 4K
>>>> content as request. The code actually is a client code which connects
>>> out to
>>>> a backend http server.
>>>>
>>>> I am using Netty 3.1.0.ALPHA2 .
>>>>
>>>> ====================
>>>> NioClientSocketChannel(id: 22f23fe5-011e-1000-ab7e-017cbfed9cbb) -
>>> (cause:
>>>> NullPointerException)
>>>> java.lang.NullPointerException
>>>> at
>>>> sun.nio.ch.EPollSelectorImpl.implRegister(EPollSelectorImpl.java:144)
>>>> at sun.nio.ch.SelectorImpl.register(SelectorImpl.java:115)
>>>> at
>>>>
>>> java.nio.channels.spi.AbstractSelectableChannel.register(AbstractSelectableChannel.java:180)
>>>
>>>> at
>>>>
>>> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.register(NioClientSocketPipelineSink.java:211)
>>>
>>>> at
>>>>
>>> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink.connect(NioClientSocketPipelineSink.java:154)
>>>
>>>> at
>>>>
>>> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink.eventSunk(NioClientSocketPipelineSink.java:106)
>>>
>>>> at org.jboss.netty.channel.Channels.connect(Channels.java:554)
>>>> at
>>>>
>>> org.jboss.netty.channel.AbstractChannel.connect(AbstractChannel.java:166)
>>>> at
>>>>
>>> org.jboss.netty.bootstrap.ClientBootstrap$Connector.channelOpen(ClientBootstrap.java:269)
>>>
>>>> at
>>>> org.jboss.netty.channel.Channels.fireChannelOpen(Channels.java:189)
>>>> at
>>>>
>>> org.jboss.netty.channel.socket.nio.NioClientSocketChannel.<init>(NioClientSocketChannel.java:88)
>>>
>>>> at
>>>>
>>> org.jboss.netty.channel.socket.nio.NioClientSocketChannelFactory.newChannel(NioClientSocketChannelFactory.java:145)
>>>
>>>> at
>>>>
>>> org.jboss.netty.channel.socket.nio.NioClientSocketChannelFactory.newChannel(NioClientSocketChannelFactory.java:92)
>>>
>>>> at
>>>>
>>> org.jboss.netty.bootstrap.ClientBootstrap.connect(ClientBootstrap.java:224)
>>>> at
>>>>
>>> org.jboss.netty.bootstrap.ClientBootstrap.connect(ClientBootstrap.java:190)
>>>> at
>>>>
>>> com.acme.test.http.netty.ia.InvocationAgentImpl.handleRequest(InvocationAgentImpl.java:127)
>>>
>>>> at
>>>>
>>> com.acme.runtime.ep.ia.impl.AsynchronousRequestProcessor$3.doInContext(AsynchronousRequestProcessor.java:186)
>>>
>>>> at
>>>>
>>> com.acme.runtime.ep.impl.MessageContextImpl.doInUserContext(MessageContextImpl.java:92)
>>>
>>>> at
>>>>
>>> com.acme.runtime.ep.ia.impl.AsynchronousRequestProcessor.doProcess(AsynchronousRequestProcessor.java:181)
>>>
>>>> at
>>>>
>>> com.acme.runtime.ep.ia.impl.RequestProcessor$1.doInContext(RequestProcessor.java:255)
>>>
>>>> at
>>>>
>>> com.acme.runtime.ep.impl.MessageContextImpl.doInContext(MessageContextImpl.java:70)
>>>
>>>> at
>>>>
>>> com.acme.runtime.ep.ia.impl.RequestProcessor.doProcess(RequestProcessor.java:206)
>>>
>>>> at
>>>>
>>> com.acme.runtime.ep.ia.impl.RequestProcessor.doProcess(RequestProcessor.java:60)
>>>
>>>> at
>>>>
>>> com.acme.runtime.ep.impl.seda.BaseProcessor.process(BaseProcessor.java:89)
>>>> at com.acme.tc.seda.Stage$Worker.run(Stage.java:249)
>>>> at
>>>>
>>> com.acme.tc.seda.LinkedListExecutor$Worker.run(LinkedListExecutor.java:300)
>>>> at java.lang.Thread.run(Thread.java:619)
>>>>
>>>> The code snip I used:
>>>> =======================
>>>> NioClientSocketChannelFactory factory = new
>>>> NioClientSocketChannelFactory(Executors.newCachedThreadPool(), Executors
>>>
>>>> .newCachedThreadPool());
>>>> ClientBootstrap client = new ClientBootstrap(factory);
>>>> client.setPipelineFactory(new PipelineFactory(getComponentContext()));
>>>> client.setOption("tcpNoDelay", true);
>>>> client.setOption("keepAlive", true);
>>>>
>>>> final String host = StringUtils.trimToNull(getConfig().get("hostName"));
>>>
>>>> if (host == null)
>>>> {
>>>> getLogger().error("hostname property required");
>>>> throw new Exception("hostName property required");
>>>> }
>>>> final String port = StringUtils.trimToNull(getConfig().get("port"));
>>>> if (host == null)
>>>> {
>>>> getLogger().error("Port property required");
>>>> throw new Exception("port property required");
>>>> }
>>>>
>>>> SocketAddress staticAddress = new InetSocketAddress(host,
>>>> Integer.parseInt(port));
>>>>
>>>> ...
>>>>
>>>> ChannelFuture future = client.connect(staticAddress); // This is the
>>> line
>>>> where gets the NPE, it is the source referred above as:
>>>> InvocationAgentImpl.java:127
>>>>
>>>> future.addListener(new ChannelFutureListener()
>>>> {
>>>> @Override
>>>> public void operationComplete(ChannelFuture future) throws Exception
>>>> {
>>>> if (!future.isSuccess())
>>>> {
>>>> getLogger().error(future.getCause().toString());
>>>> // etc ... XXXXX ...
>>>> }
>>>> else
>>>> {
>>>> Channel c = future.getChannel();
>>>> c.write(http);
>>>> }
>>>> }
>>>> });
>>>>
>>>> // Also after each message
>>>> // after we send the response, we close the channel associated with this
>>>
>>>> event.
>>>> event.getChannel().close();
>>>> in the callback function: void messageReceived(final
>>> ChannelHandlerContext
>>>> chctx, final MessageEvent event)
>>>>
>>>
>>> --
>>> Trustin Lee, Principal Software Engineer, JBoss, a division of Red Hat
>>> --
>>> what we call human nature in actuality is human habit
>>> --
>>> http://gleamynode.net/
>>>
>>>
>>
>

attachment0 (204 bytes) Download Attachment

Re: Exception under heavy load when doing Boss.register.

2008-12-11T14:34:38Z

Updated to the newest trunk (rev 608), punched with 500,000 connections, and the problem is gone.

It is a nice fix, and thanks a lot for the work.

On Thu, Dec 11, 2008 at 4:23 PM, wesly smith0 <weslysmith0@...> wrote:

I will follow up when I try the newest (rev 608 today) version on svn.

On Thu, Dec 11, 2008 at 12:40 AM, Trustin Lee-2 (via Nabble) <ml-user%2B63047-178166325@...> wrote:

Could you try the latest revision in the Subversion repository? I have
made significant improvement in NioClientSocketPipelineSink.

HTH,
Trustin

On Thu, Dec 11, 2008 at 07:32:55AM +0900, Wesly smith wrote:

>
> Exception under heavy load when doing Boss.register, the stack trace is the
> below.
>
> I saw those NPE under the load > 1500 connections/seconds for about 4K
> content as request. The code actually is a client code which connects out to
> a backend http server.
>
> I am using Netty 3.1.0.ALPHA2 .
>
> ====================
> NioClientSocketChannel(id: 22f23fe5-011e-1000-ab7e-017cbfed9cbb) - (cause:
> NullPointerException)
> java.lang.NullPointerException
> at
> sun.nio.ch.EPollSelectorImpl.implRegister(EPollSelectorImpl.java:144)
> at sun.nio.ch.SelectorImpl.register(SelectorImpl.java:115)
> at
> java.nio.channels.spi.AbstractSelectableChannel.register(AbstractSelectableChannel.java:180)
> at
> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.register(NioClientSocketPipelineSink.java:211)
> at
> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink.connect(NioClientSocketPipelineSink.java:154)
> at
> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink.eventSunk(NioClientSocketPipelineSink.java:106)
> at org.jboss.netty.channel.Channels.connect(Channels.java:554)
> at
> org.jboss.netty.channel.AbstractChannel.connect(AbstractChannel.java:166)
> at
> org.jboss.netty.bootstrap.ClientBootstrap$Connector.channelOpen(ClientBootstrap.java:269)
> at
> org.jboss.netty.channel.Channels.fireChannelOpen(Channels.java:189)
> at
> org.jboss.netty.channel.socket.nio.NioClientSocketChannel.<init>(NioClientSocketChannel.java:88)
> at
> org.jboss.netty.channel.socket.nio.NioClientSocketChannelFactory.newChannel(NioClientSocketChannelFactory.java:145)
> at
> org.jboss.netty.channel.socket.nio.NioClientSocketChannelFactory.newChannel(NioClientSocketChannelFactory.java:92)
> at
> org.jboss.netty.bootstrap.ClientBootstrap.connect(ClientBootstrap.java:224)
> at
> org.jboss.netty.bootstrap.ClientBootstrap.connect(ClientBootstrap.java:190)
> at
> com.acme.test.http.netty.ia.InvocationAgentImpl.handleRequest(InvocationAgentImpl.java:127)
> at
> com.acme.runtime.ep.ia.impl.AsynchronousRequestProcessor$3.doInContext(AsynchronousRequestProcessor.java:186)
> at
> com.acme.runtime.ep.impl.MessageContextImpl.doInUserContext(MessageContextImpl.java:92)
> at
> com.acme.runtime.ep.ia.impl.AsynchronousRequestProcessor.doProcess(AsynchronousRequestProcessor.java:181)
> at
> com.acme.runtime.ep.ia.impl.RequestProcessor$1.doInContext(RequestProcessor.java:255)
> at
> com.acme.runtime.ep.impl.MessageContextImpl.doInContext(MessageContextImpl.java:70)
> at
> com.acme.runtime.ep.ia.impl.RequestProcessor.doProcess(RequestProcessor.java:206)
> at
> com.acme.runtime.ep.ia.impl.RequestProcessor.doProcess(RequestProcessor.java:60)
> at
> com.acme.runtime.ep.impl.seda.BaseProcessor.process(BaseProcessor.java:89)
> at com.acme.tc.seda.Stage$Worker.run(Stage.java:249)
> at
> com.acme.tc.seda.LinkedListExecutor$Worker.run(LinkedListExecutor.java:300)
> at java.lang.Thread.run(Thread.java:619)
>
> The code snip I used:
> =======================
> NioClientSocketChannelFactory factory = new
> NioClientSocketChannelFactory(Executors.newCachedThreadPool(), Executors
> .newCachedThreadPool());
> ClientBootstrap client = new ClientBootstrap(factory);
> client.setPipelineFactory(new PipelineFactory(getComponentContext()));
> client.setOption("tcpNoDelay", true);
> client.setOption("keepAlive", true);
>
> final String host = StringUtils.trimToNull(getConfig().get("hostName"));
> if (host == null)
> {
> getLogger().error("hostname property required");
> throw new Exception("hostName property required");
> }
> final String port = StringUtils.trimToNull(getConfig().get("port"));
> if (host == null)
> {
> getLogger().error("Port property required");
> throw new Exception("port property required");
> }
>
> SocketAddress staticAddress = new InetSocketAddress(host,
> Integer.parseInt(port));
>
> ...
>
> ChannelFuture future = client.connect(staticAddress); // This is the line
> where gets the NPE, it is the source referred above as:
> InvocationAgentImpl.java:127
>
> future.addListener(new ChannelFutureListener()
> {
> @Override
> public void operationComplete(ChannelFuture future) throws Exception
> {
> if (!future.isSuccess())
> {
> getLogger().error(future.getCause().toString());
> // etc ... XXXXX ...
> }
> else
> {
> Channel c = future.getChannel();
> c.write(http);
> }
> }
> });
>
> // Also after each message
> // after we send the response, we close the channel associated with this
> event.
> event.getChannel().close();
> in the callback function: void messageReceived(final ChannelHandlerContext
> chctx, final MessageEvent event)
>

--
Trustin Lee, Principal Software Engineer, JBoss, a division of Red Hat
--
what we call human nature in actuality is human habit
--
http://gleamynode.net/

_______________________________________________
netty-dev mailing list
netty-dev@...
https://lists.jboss.org/mailman/listinfo/netty-dev

attachment0 (204 bytes) Download Attachment