Hi André, Many thanks for all the help and suggestions, I'll try to further dig in and see if I can find a breakthrough.
On Fri, Jul 19, 2013 at 12:53 PM, André Warnier <a...@ice-sa.com> wrote: > Shariq wrote: > >> On Thu, Jul 18, 2013 at 2:59 PM, André Warnier <a...@ice-sa.com> wrote: >> >> Shariq wrote: >>> ... >>> >>> >>> Can I please get some more help with this, as I'm not sure how to >>>> proceed >>>> further?? >>>> >>>> Shariq, >>>> >>> I note that you say please, but you've already got 3 answers, 2 of them >>> pretty detailed, in just over 24 hours since your initial post. Now you >>> are asking people to go dig into the code and give you precise answers on >>> specific lines of it. >>> Tomcat is free and open-source, and the people on this list are >>> volunteers >>> who generally have a $ job which takes precedence. A bit of patience >>> thus. >>> >>> >> Understood and thanks for all the help! I have dug deep into the source my >> self and was hoping to get some clarification on my approach, I am pretty >> familiar with the code base now so thoughts that may pop up to investigate >> further is just fine. >> >> >> Maybe you are not getting the level of interest that you expect, because >>> people feel like you are trying to redo things that have already been >>> done >>> ? or maybe because you seem to ignore the advice that you've already been >>> given (memo : for your stated goal, use the AccessLogValve and the log it >>> produces; process the log and you'll get what you are after, without >>> overly >>> affecting the very operations which you are trying to measure). >>> Or else explain convincingly why that would not work for you. >>> >> >> >> Alright, the reason is that we've built a activity monitoring system on >> top >> of cassandra, hadoop & hive, and want to reuse the same for all analytics >> related webapps too. I understand that we can use a tool that process >> logs, >> but that means we need to bring in more external tools, plus we are >> extracting more data for our analytics stuff, in the valve we've written. >> >> On the other hand AccessLogValve is a valve, in my understanding we should >> be able to write another valve and extract stuff similar to the way its >> done in ALValve. However in the custom valve we've written the output are >> different, hence my query as to the inconsistent behavior ... >> >> Ok, understood. Unfortunately, I cannot help you further myself, but > hopefully this explanation will motivate more people to try to help. > Many thanks for all the help and suggestions, and yes I hope so too :) > > Note that a Valve is something specific to Tomcat. If you want > portability, then you should rather look at servlet filters. A servlet > filter "wraps" the webapp, so it is there both on the request and the > response side, and it should be able to collect the information you're > after. > With the previous caveats though : it will measure the times and data > volumes around your webapps, but not really the "bandwidth" (except in a > very broad sense). I'll try this as well, thanks .! > > > ------------------------------**------------------------------**--------- > To unsubscribe, e-mail: > users-unsubscribe@tomcat.**apache.org<users-unsubscr...@tomcat.apache.org> > For additional commands, e-mail: users-h...@tomcat.apache.org > >
