Maybe not that hard but the old system I was using did have a complex mechanism where multiple services threads are called for the whole statistics gathering process. That's why I am a bit confused. A simple implementation would be enough. Sorry if I bothered you. Thanks again.
Emre -----Original Message----- From: Alexander Christian [mailto:[email protected]] Sent: Friday, January 24, 2014 10:48 AM To: [email protected] Subject: Re: IoServiceStatistics vs StatCollector Is it really THAT hard to adapt the new and self-explaining IoSeviceStatistics class to the old StatCollector interface you where using?! Or what kind of implementation do you search?! br, Alex Am 24.01.2014 09:41, schrieb Emre Baykal: > Thanks a lot again. That would help. > But still, is there any implementations I can find on the web? I crawled the > web but can't find anything related this issue. Has anyone worked on this > before? > > Emre > > > -----Original Message----- > From: Emmanuel Lécharny [mailto:[email protected]] > Sent: Friday, January 24, 2014 10:01 AM > To: [email protected] > Subject: Re: IoServiceStatistics vs StatCollector > > Le 1/24/14 8:14 AM, Emre Baykal a écrit : >> Thanks! >> What I would like to know is the way it collects statistics. Apparently >> there is no start() or stop() method in order to start worker thread. Could >> you provide me an example? > We use the session's thread to update those statistics. For instance: > > public void fireMessageSent(WriteRequest request) { > session.increaseWrittenMessages(request, > System.currentTimeMillis()); > ... > > We then do not need a dedicated thread to manage those statistics. > >
