2014-06-17 17:08 GMT+04:00 Mark Thomas <ma...@apache.org>: > On 17/06/2014 00:50, Konstantin Kolinko wrote: >> 2014-06-14 0:21 GMT+04:00 <ma...@apache.org>: >>> Author: markt >>> Date: Fri Jun 13 20:21:32 2014 >>> New Revision: 1602510
> <snip/> >> 2) The submission of SocketProcessor to the Executor may end with >> RejectedExecutionException. >> - in NioEndpoint#processSocket(NioChannel, SocketStatus, boolean) >> >> In that case it just logs a warning and does nothing else. >> >> How likely is this rejected execution? > > Very unlikely. > >> My understanding/answer is that the rejection is unlikely. >> The executor here is o.a.c.core.StandardThreadExecutor. Its execute() >> method is unlikely to reject tasks. There is a queue that accepts >> tasks. Tasks will be rejected if either >> a) Executor is not running. >> b) Queue is full. The maximum queue length is Integer.MAX_VALUE by >> default. Though its size is configurable via >> StandardThreadExecutor.setMaxQueueSize(). >> >> Will it result in failing to decrement the connection counter? >> Will it result in failing to obtain and to recycle the processor >> associated with the socket? >> >> I suspect that per this BZ 56518 it will fail to decrement the counter. > > The only time I think this might occur is during connector stop() in > which case the socket will be closed and the connection counter reset > anyway. > ACK. Maybe I'll add a warning to the docs to discourage changing maxQueueSize. <snip/> .... > > If you have a scenario where you think this is an issue, please share it. > >> 4) There are no tests for this new feature. > > I have been using the test WAR provided by the OP. I haven't had a > chance to convert it to a unit test. Obviously, no objections from me if > you want to do that. Ack. Thanks for the pointer. (I am not sure that I'll find the time, but now I know where to look). In Tomcat 8 there was the following commit as part of fixing this issue: http://svn.apache.org/viewvc?view=revision&revision=r1594411 Are there plans to backport it to Tomcat 7? Best regards, Konstantin Kolinko --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
