Fix doc
This commit is contained in:
Frédéric Brégier
parent
eb4e6bd556
commit
66f645cd55
@ -28,21 +28,22 @@
|
||||
*
|
||||
* <P>The main goal of this package is to allow to shape the traffic (bandwidth limitation),
|
||||
* but also to get statistics on how many bytes are read or written. Both functions can
|
||||
* be active or inactive (traffic or statistics).</P><br><br><br>
|
||||
* be active or inactive (traffic or statistics).</P>
|
||||
*
|
||||
* <P>Three classes implement this behavior:<br><br>
|
||||
* <P>Three classes implement this behavior:<br>
|
||||
* <ul>
|
||||
* <li>) <tt>PerformanceCounter</tt>: this class is the kernel of the package. It can be accessed to get some extra information
|
||||
* <li> <tt>PerformanceCounter</tt>: this class is the kernel of the package. It can be accessed to get some extra information
|
||||
* like the read or write bytes since last check, the read and write bandwidth from last check...</li><br><br>
|
||||
*
|
||||
* <li>) <tt>PerformanceCounterFactory</tt>: this class has to be implemented in your code in order to implement (eventually empty)
|
||||
* <li> <tt>PerformanceCounterFactory</tt>: this class has to be implemented in your code in order to implement (eventually empty)
|
||||
* the accounting method. This class is a Factory for PerformanceCounter which is used in the third class to create the
|
||||
* necessary PerformanceCounter according to your specifications.</li><br><br>
|
||||
*
|
||||
* <li>) <tt>TrafficShapingHandler</tt>: this class is the handler to be inserted in your pipeline. The insertion can be wherever
|
||||
* you want, but <b>it must be placed before any <tt>MemoryAwareThreadPoolExecutor</tt></b> in your pipeline.</li><br><br>
|
||||
* <li> <tt>TrafficShapingHandler</tt>: this class is the handler to be inserted in your pipeline. The insertion can be wherever
|
||||
* you want, but <b>it must be placed before any <tt>MemoryAwareThreadPoolExecutor</tt> in your pipeline</b>.</li><br>
|
||||
* <b><i>It is really recommended
|
||||
* to have such a <tt>emoryAwareThreadPoolExecutor</tt> (either non ordered or <tt>OrderedMemoryAwareThreadPoolExecutor</tt>) in your pipeline</i></b>
|
||||
* to have such a</i> <tt>MemoryAwareThreadPoolExecutor</tt> <i>(either non ordered or </i><tt>OrderedMemoryAwareThreadPoolExecutor</tt>
|
||||
* <i>) in your pipeline</i></b>
|
||||
* when you want to use this feature with some real traffic shaping, since it will allow to relax the constraint on
|
||||
* NioWorker to do other jobs if necessary.<br>
|
||||
* Instead, if you don't, you can have the following situation: if there are more clients
|
||||
@ -53,39 +54,41 @@
|
||||
* stopping by this handler.<br><br>
|
||||
* The method <tt>getMessageSize(MessageEvent)</tt> has to be implemented to specify what is the size of the object to be read or write
|
||||
* accordingly to the type of object. In simple case, it can be as simple as a call to <tt>getChannelBufferMessageSize(MessageEvent)</tt>.<br>
|
||||
* </ul></P>
|
||||
*
|
||||
* <P>Standard use could be as follow:</P>
|
||||
*
|
||||
* <P><ul>
|
||||
* <li>To activate or deactivate the traffic shaping, change the value corresponding to your desire as
|
||||
* [Global or per Channel] [Write or Read] Limitation in byte/s.</li><br>
|
||||
* <tt>PerformanceCounterFactory.NO_LIMIT</tt> (-1)
|
||||
* stands for no limitation, so the traffic shaping is deactivate (on what you specified).<br>
|
||||
* You can either change those values with the method <tt>changeConfiguration</tt> in PerformanceCounterFactory or
|
||||
* directly from the PerformanceCounter method <tt>changeConfiguration</tt>.<br>
|
||||
* <br>
|
||||
*
|
||||
* <li>To activate or deactivate the statistics, you can adjust the delay to a low (not less than 200ms
|
||||
* for efficiency reasons) or a high value (let say 24H in ms is huge enough to not get the problem)
|
||||
* or even using <tt>PerformanceCounterFactory.NO_STAT</tt> (-1)</li>.<br>
|
||||
* And if you don't want to do anything with this statistics, just implement an empty method for
|
||||
* <tt>PerformanceCounterFactory.accounting(PerformanceCounter)</tt>.<br>
|
||||
* Again this can be changed either from PerformanceCounterFactory or directly in PerformanceCounter.<br><br>
|
||||
*
|
||||
* <li>You can also completely deactivate channel or global PerformanceCounter by setting the boolean to false
|
||||
* accordingly to your needs in the PerformanceCounterFactory. It will deactivate the global Monitor. For channels monitor,
|
||||
* it will prevent new monitors to be created (or reversely they will be created for newly connected channels).</li>
|
||||
* </ul></P><br><br>
|
||||
*
|
||||
* <P>Standard use could be as follow:</P><br><br>
|
||||
*
|
||||
* <P>To activate or deactivate the traffic shaping, change the value corresponding to your desire as
|
||||
* [Global or per Channel] [Write or Read] Limitation in byte/s.</P><br>
|
||||
* <P><tt>PerformanceCounterFactory.NO_LIMIT</tt> (-1)
|
||||
* stands for no limitation, so the traffic shaping is deactivate (on what you specified).</P><br>
|
||||
* <P>You can either change those values with the method <tt>changeConfiguration</tt> in PerformanceCounterFactory or
|
||||
* directly from the PerformanceCounter method <tt>changeConfiguration</tt>.</P><br>
|
||||
* <br><br>
|
||||
*
|
||||
* <P>To activate or deactivate the statistics, you can adjust the delay to a low (not less than 200ms
|
||||
* for efficiency reasons) or a high value (let say 24H in ms is huge enough to not get the problem)
|
||||
* or even using <tt>PerformanceCounterFactory.NO_STAT</tt> (-1).</P><br>
|
||||
* <P>And if you don't want to do anything with this statistics, just implement an empty method for
|
||||
* <tt>PerformanceCounterFactory.accounting(PerformanceCounter)</tt>.</P><br>
|
||||
* <P>Again this can be changed either from PerformanceCounterFactory or directly in PerformanceCounter.</P><br><br><br>
|
||||
*
|
||||
* <P>You can also completely deactivate channel or global PerformanceCounter by setting the boolean to false
|
||||
* accordingly to your needs in the PerformanceCounterFactory. It will deactivate the global Monitor. For channels monitor,
|
||||
* it will prevent new monitors to be created (or reversely they will be created for newly connected channels).</P><br><br><br>
|
||||
*
|
||||
* <P>So in your application you will create your own PerformanceCounterFactory and setting the values to fit your needs.</P><br><br>
|
||||
* <P>So in your application you will create your own PerformanceCounterFactory and setting the values to fit your needs.</P>
|
||||
* <tt>MyPerformanceCounterFactory myFactory = new MyPerformanceCounter(...);</tt><br><br><br>
|
||||
* <P>Then you can create your pipeline (using a PipelineFactory since each TrafficShapingHandler must be unique by channel) and adding this handler before
|
||||
* your MemoryAwareThreadPoolExecutor:</P><br><br>
|
||||
* your MemoryAwareThreadPoolExecutor:</P>
|
||||
* <tt>pipeline.addLast("MyTrafficShaping",new MyTrafficShapingHandler(myFactory));</tt><br>
|
||||
* <tt>...</tt><br>
|
||||
* <tt>pipeline.addLast("MemoryExecutor",new ExecutionHandler(memoryAwareThreadPoolExecutor));</tt><br><br><br>
|
||||
* <tt>pipeline.addLast("MemoryExecutor",new ExecutionHandler(memoryAwareThreadPoolExecutor));</tt><br><br>
|
||||
*
|
||||
* <P>TrafficShapingHandler must be unique by channel but however it is still global due to
|
||||
* the PerformanceCounterFactcory that is shared between all handlers across the channels.</P><br><br>
|
||||
* the PerformanceCounterFactcory that is shared between all handlers across the channels.</P>
|
||||
*
|
||||
*
|
||||
*
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user