Class FlushConsolidationHandler

java.lang.Object
io.netty.channel.ChannelHandlerAdapter
All Implemented Interfaces:
ChannelHandler, ChannelInboundHandler, ChannelOutboundHandler
public class FlushConsolidationHandler extends ChannelDuplexHandler
ChannelDuplexHandler which consolidates Channel.flush() / ChannelHandlerContext.flush() operations (which also includes Channel.writeAndFlush(Object) / Channel.writeAndFlush(Object, ChannelPromise) and ChannelOutboundInvoker.writeAndFlush(Object) / ChannelOutboundInvoker.writeAndFlush(Object, ChannelPromise)).

Flush operations are generally speaking expensive as these may trigger a syscall on the transport level. Thus it is in most cases (where write latency can be traded with throughput) a good idea to try to minimize flush operations as much as possible.

If a read loop is currently ongoing, flush(ChannelHandlerContext) will not be passed on to the next ChannelOutboundHandler in the ChannelPipeline, as it will pick up any pending flushes when channelReadComplete(ChannelHandlerContext) is triggered. If no read loop is ongoing, the behavior depends on the consolidateWhenNoReadInProgress constructor argument:

  • if false, flushes are passed on to the next handler directly;
  • if true, the invocation of the next handler is submitted as a separate task on the event loop. Under high throughput, this gives the opportunity to process other flushes before the task gets executed, thus batching multiple flushes into one.
If explicitFlushAfterFlushes is reached the flush will be forwarded as well (whether while in a read loop, or while batching outside of a read loop).

If the Channel becomes non-writable it will also try to execute any pending flush operations.

The FlushConsolidationHandler should be put as first ChannelHandler in the ChannelPipeline to have the best effect.