TC
<!-- Uncomment this to enable Comet connection tacking (provides events
on session expiration as well as webapp lifecycle) -->
<!--
<Valve className="org.apache.catalina.valves.CometConnectionManagerValve" />
-->
<property name="org.apache.tomcat.util. net.NioSelectorShared">
<p>If <code>true</code>, use a shared selector for servlet write/read. If
not specified, the default value of <code>true</code> will be used.</p>
</property>
<code>org.apache.coyote.http11.Http11NioProtocol</code> - non blocking Java
connector<br/>
<!-- configurable attributes are -->
<subsection name="Nio Implementation">
<p>The NIO connector exposes all the low level socket properties that can
be used to tune the connector.
Most of these attributes are directly linked to the socket implementation
in the JDK so you can find out
about the actual meaning in the JDK API documentation.<br/>
<strong>Note</strong>On some JDK versions, setTrafficClass causes a
problem, a work around for this is to add
the <code>-Djava.net.preferIPv4Stack=true</code> value to your command
line</p>
<attributes>
<attribute name="useSendfile" required="false">
<p>(bool)Use this attribute to enable or disable sendfile capability.
The default value is <code>true</code>
</p>
</attribute>
<attribute name="useExecutor" required="false">
<p>(bool)Set to true to use the NIO thread pool executor. The default
value is <code>true</code>.
If set to false, it uses a thread pool based on a stack for its
execution.
Generally, using the executor yields a little bit slower performance,
but yields a better
fairness for processing connections in a high load environment as the
traffic gets queued through a
FIFO queue. If set to true(default) then the max pool size is the
<code>maxThreads</code> attribute
and the core pool size is the <code>minSpareThreads</code>.
This value is ignored if the <code>executor</code> attribute is present
and points to a valid shared thread pool.
</p>
</attribute>
<attribute name="executor" required="false">
<p>A reference to the name in an <a href="executor.html">Executor</a>
element.
If this attribute is enabled, and the named executor exists, the
connector will
use the executor, and all the other thread attributes will be
ignored.</p>
</attribute>
<attribute name="acceptorThreadCount" required="false">
<p>(int)The number of threads to be used to accept connections.
Increase this value on a multi CPU machine,
although you would never really need more than <code>2</code>. Also,
with a lot of non keep alive connections,
you might want to increase this value as well. Default value is
<code>1</code>.</p>
</attribute>
<attribute name="pollerThreadCount" required="false">
<p>(int)The number of threads to be used to run for the polling events.
Default value is <code>1</code> per processor. Can't see a reason to go
above that. But experiment and find your own results.</p>
</attribute>
<attribute name="pollerThreadPriority" required="false">
<p>(int)The priority of the poller threads.
The default value is <code>java.lang.Thread#NORM_PRIORITY</code>.
See the JavaDoc for the java.lang.Thread class for more details on
what this priority means.
</p>
</attribute>
<attribute name="acceptorThreadPriority" required="false">
<p>(int)The priority of the acceptor threads. The threads used to
accept new connections.
The default value is <code>java.lang.Thread#NORM_PRIORITY</code>.
See the JavaDoc for the java.lang.Thread class for more details on
what this priority means.
</p>
</attribute>
<attribute name="selectorTimeout" required="false">
<p>(int)The time in milliseconds to timeout on a select() for the
poller.
This value is important, since connection clean up is done on the
same thread, so dont set this
value to an extremely high one. The default value is
<code>1000</code> milliseconds.</p>
</attribute>
<attribute name="useComet" required="false">
<p>(bool)Whether to allow comet servlets or not, Default value is
<code>true</code>.</p>
</attribute>
<attribute name="processCache" required="false">
<p>(int)The protocol handler caches Http11NioProcessor objects to speed
up performance.
This setting dictates how many of these objects get cached.
<code>-1</code> means unlimited, default is <code>200</code>. Set
this value somewhere close to your maxThreads value.
</p>
</attribute>
<attribute name="maxKeepAliveRequests" required="false">
<p>The maximum number of HTTP requests which can be pipelined until
the connection is closed by the server. Setting this attribute to 1 will
disable HTTP/1.0 keep-alive, as well as HTTP/1.1 keep-alive and
pipelining. Setting this to -1 will allow an unlimited amount of
pipelined or keep-alive HTTP requests.
If not specified, this attribute is set to 100.</p>
</attribute>
<attribute name="socket.directBuffer" required="false">
<p>(bool)Boolean value, whether to use direct ByteBuffers or java
mapped ByteBuffers. Default is <code>false</code>
<br/>When you are using direct buffers, make sure you allocate the
appropriate amount of memory for the
direct memory space. On Sun's JDK that would be something like
<code>-XX:MaxDirectMemorySize=256m</code></p>
</attribute>
<attribute name="socket.rxBufSize" required="false">
<p>(int)The socket receive buffer (SO_RCVBUF) size in bytes. Default
value is <code>25188</code></p>
</attribute>
<attribute name="socket.txBufSize" required="false">
<p>(int)The socket send buffer (SO_SNDBUF) size in bytes. Default value
is <code>43800</code></p>
</attribute>
<attribute name="socket.appReadBufSize" required="false">
<p>(int)Each connection that is opened up in Tomcat get associated with
a read and a write ByteBuffer
This attribute controls the size of these buffers. By default this
read buffer is sized at <code>8192</code> bytes.
For lower concurrency, you can increase this to buffer more data.
For an extreme amount of keep alive connections, decrease this
number or increase your heap size.</p>
</attribute>
<attribute name="socket.appWriteBufSize" required="false">
<p>(int)Each connection that is opened up in Tomcat get associated with
a read and a write ByteBuffer
This attribute controls the size of these buffers. By default this
write buffer is sized at <code>8192</code> bytes.
For low concurrency you can increase this to buffer more response
data.
For an extreme amount of keep alive connections, decrease this
number or increase your heap size.
<br/>
The default value here is pretty low, you should up it if you are
not dealing with tens of thousands
concurrent connections.</p>
</attribute>
<attribute name="socket.bufferPool" required="false">
<p>(int)The Nio connector uses a class called NioChannel that holds
elements linked to a socket.
To reduce garbage collection, the Nio connector caches these channel
objects.
This value specifies the size of this cache.
The default value is <code>500</code>, and represents that the cache
will hold 500 NioChannel objects.
Other values are <code>-1</code>. unlimited cache, and
<code>0</code>, no cache.</p>
</attribute>
<attribute name="socket.bufferPoolSize" required="false">
<p>(int)The NioChannel pool can also be size based, not used object
based. The size is calculated as follows:<br/>
NioChannel <code>buffer size = read buffer size + write buffer
size</code><br/>
SecureNioChannel <code>buffer size = application read buffer size +
application write buffer size + network read buffer size + network write buffer
size</code><br/>
The value is in bytes, the default value is <code>1024*1024*100</code>
(100MB)
</p>
</attribute>
<attribute name="socket.processorCache" required="false">
<p>(int)Tomcat will cache SocketProcessor objects to reduce garbage
collection.
The integer value specifies how many objects to keep in the cache at
most.
The default is <code>500</code>.
Other values are <code>-1</code>. unlimited cache, and
<code>0</code>, no cache.</p>
</attribute>
<attribute name="socket.keyCache" required="false">
<p>(int)Tomcat will cache KeyAttachment objects to reduce garbage
collection.
The integer value specifies how many objects to keep in the cache at
most.
The default is <code>500</code>.
Other values are <code>-1</code>. unlimited cache, and
<code>0</code>, no cache.</p>
</attribute>
<attribute name="socket.eventCache" required="false">
<p>(int)Tomcat will cache PollerEvent objects to reduce garbage
collection.
The integer value specifies how many objects to keep in the cache at
most.
The default is <code>500</code>.
Other values are <code>-1</code>. unlimited cache, and
<code>0</code>, no cache.</p>
</attribute>
<attribute name="socket.tcpNoDelay" required="false">
<p>(bool)same as the standard setting <code>tcpNoDelay</code>. Default
value is <code>false</code></p>
</attribute>
<attribute name="socket.soKeepAlive" required="false">
<p>(bool)Boolean value for the socket's keep alive setting
(SO_KEEPALIVE). Default is <code>false</code>. </p>
</attribute>
<attribute name="socket.ooBInline" required="false">
<p>(bool)Boolean value for the socket OOBINLINE setting. Default value
is <code>true</code></p>
</attribute>
<attribute name="socket.soReuseAddress" required="false">
<p>(bool)Boolean value for the sockets reuse address option
(SO_REUSEADDR). Default value is <code>true</code></p>
</attribute>
<attribute name="socket.soLingerOn" required="false">
<p>(bool)Boolean value for the sockets so linger option (SO_LINGER).
Default value is <code>true</code>.
This option is paired with the <code>soLingerTime</code> value.</p>
</attribute>
<attribute name="socket.soLingerTime" required="false">
<p>(bool)Value in seconds for the sockets so linger option (SO_LINGER).
Default value is <code>25</code> seconds.
This option is paired with the soLinger value.</p>
</attribute>
<attribute name="socket.soTimeout" required="false">
<p>(int)Value in milliseconds for the sockets read timeout
(SO_TIMEOUT). Default value is <code>5000</code> milliseconds.</p>
</attribute>
<attribute name="socket.soTrafficClass" required="false">
<p>(byte)Value between <code>0</code> and <code>255</code> for the
traffic class on the socket, <code>0x04 | 0x08 | 0x010</code></p>
</attribute>
<attribute name="socket.performanceConnectionTime" required="false">
<p>(int)The first value for the performance settings. Default is
<code>1</code>, see <a
href="http://java.sun.com/j2se/1.5.0/docs/api/java/net/Socket.html#setPerformancePreferences(int,%20int,%20int)">Socket
Performance Options</a></p>
</attribute>
<attribute name="socket.performanceLatency" required="false">
<p>(int)The second value for the performance settings. Default is
<code>0</code>, see <a
href="http://java.sun.com/j2se/1.5.0/docs/api/java/net/Socket.html#setPerformancePreferences(int,%20int,%20int)">Socket
Performance Options</a></p>
</attribute>
<attribute name="socket.performanceBandwidth" required="false">
<p>(int)The third value for the performance settings. Default is
<code>1</code>, see <a
href="http://java.sun.com/j2se/1.5.0/docs/api/java/net/Socket.html#setPerformancePreferences(int,%20int,%20int)">Socket
Performance Options</a></p>
</attribute>
<attribute name="selectorPool.maxSelectors" required="false">
<p>(int)The max selectors to be used in the pool, to reduce selector
contention.
Use this option when the command line
<code>org.apache.tomcat.util.net.NioSelectorShared</code> value is set to false.
Default value is <code>200</code>.</p>
</attribute>
<attribute name="selectorPool.maxSpareSelectors" required="false">
<p>(int)The max spare selectors to be used in the pool, to reduce
selector contention.
When a selector is returned to the pool, the system can decide to keep
it or let it be GC:ed.
Use this option when the command line
<code>org.apache.tomcat.util.net.NioSelectorShared</code> value is set to false.
Default value is <code>-1</code> (unlimited)</p>
</attribute>
<attribute name="command-line-options" required="false">
<p>The following command line options are available for the NIO
connector:<br/>
<code>-Dorg.apache.tomcat.util.net.NioSelectorShared=true|false</code>
- default is <code>true</code>.
Set this value to false if you wish to use a selector for each thread.
the property. If you do set it to false, you can control the size of
the pool of selectors by using the
selectorPool.maxSelectors attribute</p>
</attribute>
<attribute name="oomParachute" required="false">
<p>(int)The NIO connector implements an OutOfMemoryError strategy
called parachute.
It holds a chunk of data as a byte array. In case of an OOM,
this chunk of data is released and the error is reported. This will
give the VM enough room
to clean up. The <code>oomParachute</code> represent the size in
bytes of the parachute(the byte array).
The default value is <code>1024*1024</code>(1MB).
Please note, this only works for OOM errors regarding the Java Heap
space, and there is absolutely no
guarantee that you will be able to recover at all.
If you have an OOM outside of the Java Heap, then this parachute
trick will not help.
</p>
</attribute>
</attributes>
</subsection>
<p>Below is a small chart that shows how the connectors differentiate.</p>
<source>
Java Blocking Connector Java Nio Blocking
Connector APR Connector
Classname Http11Protocol Http11NioProtocol
Http11AprProtocol
Tomcat Version 3.x 4.x 5.x 6.x 6.x
5.5.x 6.x
Support Polling NO YES
YES
Polling Size N/A Unlimited - Restricted by
mem Unlimited - Configurable
Read HTTP Request Blocking Non Blocking
Blocking
Read HTTP Body Blocking Sim Blocking
Blocking
Write HTTP Response Blocking Sim Blocking
Blocking
SSL Support Java SSL Java SSL
OpenSSL
SSL Handshake Blocking Non blocking
Blocking
Max Connections maxThreads See polling size
See polling size
you can configure the NIOConnector with attributes and valves as mentioned
above or can implement Tribes on TC which uses Tribes
http://tomcat.apache.org/tomcat-6.0-doc/tribes/introduction.html
Martin Gainty
______________________________________________
Verzicht und Vertraulichkeitanmerkung/Note de déni et de confidentialité
Diese Nachricht ist vertraulich. Sollten Sie nicht der vorgesehene Empfaenger
sein, so bitten wir hoeflich um eine Mitteilung. Jede unbefugte Weiterleitung
oder Fertigung einer Kopie ist unzulaessig. Diese Nachricht dient lediglich dem
Austausch von Informationen und entfaltet keine rechtliche Bindungswirkung.
Aufgrund der leichten Manipulierbarkeit von E-Mails koennen wir keine Haftung
fuer den Inhalt uebernehmen.
Ce message est confidentiel et peut être privilégié. Si vous n'êtes pas le
destinataire prévu, nous te demandons avec bonté que pour satisfaire informez
l'expéditeur. N'importe quelle diffusion non autorisée ou la copie de ceci est
interdite. Ce message sert à l'information seulement et n'aura pas n'importe
quel effet légalement obligatoire. Étant donné que les email peuvent facilement
être sujets à la manipulation, nous ne pouvons accepter aucune responsabilité
pour le contenu fourni.
> Date: Mon, 28 Jun 2010 01:28:15 -0700
> From: roger.var...@googlemail.com
> To: user@struts.apache.org
> Subject: Struts 2, Comet & WebSockets
>
>
> Hi
>
> Does Struts 2 have support for (or plans for supporting) either Comet
> processing or HTML5 WebSockets. Googling suggests that the answer for Comet
> may be no, but the latest page I've found that states this, is from 2008!
>
> Regards
> --
> View this message in context:
> http://old.nabble.com/Struts-2%2C-Comet---WebSockets-tp29008515p29008515.html
> Sent from the Struts - User mailing list archive at Nabble.com.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscr...@struts.apache.org
> For additional commands, e-mail: user-h...@struts.apache.org
>
_________________________________________________________________
The New Busy think 9 to 5 is a cute idea. Combine multiple calendars with
Hotmail.
http://www.windowslive.com/campaign/thenewbusy?tile=multicalendar&ocid=PID28326::T:WLMTAGL:ON:WL:en-US:WM_HMP:042010_5
