This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
which really affects performance. Since the effect is subtle and it is difficult to fully disable monitoring once it is turned on, we need to have http monitoring turned off by default. The user needs to get enough information about the effect of turning on the monitoring to allow them to make a smart choice for their environment. To resolve this issue for 6.5, we will: 1. not enable http monitoring as a default choice... 2. draft a complete description of the implementation of http monitoring and its effect on server performance 3. create a warning dialog, that will have a reference to the complete description document. 4. rewrite the http monitor support classes in the GF plugins to account for the differences between Tomcat and GF
step 4 is a time sink that may not be worth doing... The server teams has resolved this behavior differences between TC and GF... see https://glassfish.dev.java.net/issues/show_bug.cgi?id=3844. The fix is scheduled to release in December according to http://forums.java.net/jive/thread.jspa?threadID=44301&tstart=0
tested v2.1 server bits... fix for gf 3844 doesn't actually address this issue... added info to gf issue and re-opened.
part one (do not have http monitor on by default): http://hg.netbeans.org/main/rev/68b511536e6c part three (warn user about effect on caching): http://hg.netbeans.org/main/rev/f04de015d5b1