From: Kent N. <ken...@se...> - 2009-02-04 00:09:37
|
I am considering using JETM for measuring "execution" time for asynchronous request sent externally. Namely, one set of threads works on sending out requests and another set of threads processes incoming requests (we are talking very high throughput here, possibly 1000's of messages/sec). Now if I use JETM and create a point in one thread and collect it in another, will this cause a problem? I haven't seen any specific info in the docs regarding this, but I get the feeling the context and the aggregation chain of the points are deducted from the thread context, or? Also, this is a benchmarking application, so I would like to be able to get some very simplified results at the end, like: - Total messages sent/received - Average/max/min request time - Messages sent/received per second in average (ie total divided by duration) Is there a finished renderer that can be configured to show info like this easily? Or is this easiest solved by generating an RRD database and then using RRD tools to extract this information? |