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.
NB Full Dev (200706180000), JDK 1.5.0_12 & JDK 6.0_01 We are measuring regression on Startup time - compare to 200706140000 ------------- 1st usage 5% 3% 6% 6% 5% 4% 3% Subsequent usage 5% 4% 4% 4% 6% 6% 5% ------------- 1st usage 13770 30088 22563 24689 46243 75680 52712 Subsequent usage 9255 21702 13337 16818 34259 53708 35341 .... recommended HW: Ubuntu6.10 Sol10 WinXp WinVis minimum HW: Fedora3 Sol10 WinXP we are also measuring regression on Startup time with opened Java project usage 10% 5% 10% 8% ... on reccomnded HW only : Ubuntu6.10 Sol10 WinXP WinVis There is a regression on - Window System shown - 5% 4% 5% 2% 13% 6% 3% http://xtest-db.czech.sun.com/AllInOnePerformanceDashboard/atomicResult.jsp?buildNumber=200706180000&suiteName=Startup%20Time&resultName=Startup%20Time%20|%20Window%20system%20shown - ModuleSystem.readList - 3% 1% 2% 1% 1% 3% 1% http://xtest-db.czech.sun.com/AllInOnePerformanceDashboard/atomicResult.jsp?buildNumber=200706180000&suiteName=Startup%20Time&resultName=Startup%20Time%20|%20ModuleSystem.readList%20finished Other regressions on complex startup between same builds - startup with opened Visual Web project - issue 107233 - startup with opened Mobility project - issue 107224
The regression was caused by addition of CND cluster. We can't do much about it, except implementing larger scalability fixes in the module system, but an addition of a block of modules will always have some impact (not talking about badly written modules with slow ModuleInstalls). Note: CND's ModuleInstalls alone consume ~1% of the warm startup time. CND's mime resolvers have also contributed to overall slowdown of the IDE, but this was already fixed.
v/c