Hi there, This is directed at those involved in the development of the org.freeswitch.esl.client library. Apologies if this has already been mentioned, I couldn't find any references to it. In evaluating this library, I noticed that AbstractOutboundPipelineFactory implementation of ChannelPipelineFactory.getPipeline() creates a new ExecutionHandler each time it is called, which results in a increasing number of thread pools being used that eventually triggers the following warning: org.jboss.netty.util.internal.SharedResourceMisuseDetector WARNING: You are creating too many MemoryAwareThreadPoolExecutor instances. MemoryAwareThreadPoolExecutor is a shared resource that must be reused across the application, so that only a few instances are created. Looking at the documentation for ExecutionHandler, I noticed that the example implementation provides a dedicated ExecutionHandler instance, which is passed to all pipelines returned in getPipeline(). Note the comment in the example that states it "Must be shared".
From http://docs.jboss.org/netty/3.2/api/org/jboss/netty/handler/execution/Execut...:
public class DatabaseGatewayPipelineFactory implements |ChannelPipelineFactory| <http://docs.jboss.org/netty/3.2/api/org/jboss/netty/channel/ChannelPipelineFactory.html> { *private final |ExecutionHandler| <http://docs.jboss.org/netty/3.2/api/org/jboss/netty/handler/execution/ExecutionHandler.html> executionHandler;* public DatabaseGatewayPipelineFactory(|ExecutionHandler| <http://docs.jboss.org/netty/3.2/api/org/jboss/netty/handler/execution/ExecutionHandler.html> executionHandler) { this.executionHandler = executionHandler; } public |ChannelPipeline| <http://docs.jboss.org/netty/3.2/api/org/jboss/netty/channel/ChannelPipeline.html> getPipeline() { return |Channels| <http://docs.jboss.org/netty/3.2/api/org/jboss/netty/channel/Channels.html>.pipeline( new DatabaseGatewayProtocolEncoder(), new DatabaseGatewayProtocolDecoder(), * executionHandler, // Must be shared* new DatabaseQueryingHandler()); } } ------------- Is there any particular reason why AbstractOutboundPipelineFactory does not provide a dedicated ExecutionHandler that may be shared by all pipelines, or is this simply an oversight? Thanks and keep up the good work, Adam Kelloway