Mine is a J2EE web application, which I am trying to monitor with by hooking up the DT Agent 6.0.0 to my JVM.
There is one particular business transaction that manually takes more than 30 sec to completed and I am trying to figure out (with help of DT) which method calls are taking more time. i doubt that this transaction is doing lot of I/O and the method is suspended for a while, which is why a high delay.
But to my surprise, if i look at the "Web Request" page I see all the sub-calls (URLs) getting listed but the Total Sum[ms] value for all the sub-urls is just few milli-seconds. This is weird and as it doesn't correlate with what i am getting from web browser
Even, I recorded that flow in LoadRunner and ran for few iteration in Vugen (just to check the delay without rendering factor) and every time i got same 30+sec response from my particular transaction. But somehow DT is not showing such high value in the results.
has anyone faced the same issue? do i need to change my Monitoring Group profile?
Hi. Do you have a web server (Apache, IIS, NGINX, ...) in front of your J2EE Server? If that is the case you could install a web server agent in that web server to see if time is lost in the web server. We have seen this many times that time is spent in tiers BEFORE the Java App Server. Web Server or even Load Balancers could be a problem.
We had a similar problem using mod_proxy. There is a setting to increase the threads to the backend server. I have included both links to help.
The reason dynaTrace doesn't show a problem is because there isn't one. Your server is serving pages as fast as they can, however the proxy is limiting the number of concurrent.
|max||1...n||Maximum number of connections that will be allowed to the backend server. The default for this limit is the number of threads per process in the active MPM. In the Prefork MPM, this is always 1, while with other MPMs it is controlled by the |