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.
Product Version = NetBeans IDE Dev (Build 201305262300) Operating System = Windows 7 version 6.1 running on amd64 Java; VM; Vendor = 1.7.0_17 Runtime = Java HotSpot(TM) 64-Bit Server VM 23.7-b01 While editing a XML file in source code (more precisely, a Persistence Unit xml file), after the typing <!--, IDE is blocked for a long time. After is released, and I can finish the block. In IDE log I can see the following message: INFO [org.netbeans.modules.j2ee.persistence.unit.PUDataObject] java.lang.RuntimeException: DOM graph creation failed: org.netbeans.modules.schema2beans.Schema2BeansRuntimeException: Failed to create the XML-DOM Document. Check your XML to make sure it is correct. The string "--" is not permitted within comments. at org.netbeans.modules.j2ee.persistence.dd.persistence.model_2_0.Persistence.createGraph(Persistence.java:235) at org.netbeans.modules.j2ee.persistence.dd.persistence.model_2_0.Persistence.createGraph(Persistence.java:226) [catch] at org.netbeans.modules.j2ee.persistence.unit.PUDataObject.parseDocument(PUDataObject.java:215) at org.netbeans.modules.j2ee.persistence.unit.PUDataObject$ModelSynchronizer.reloadModelFromData(PUDataObject.java:500) at org.netbeans.modules.xml.multiview.XmlMultiViewDataSynchronizer.reloadModel(XmlMultiViewDataSynchronizer.java:230) at org.netbeans.modules.xml.multiview.XmlMultiViewDataSynchronizer$3.run(XmlMultiViewDataSynchronizer.java:100) at org.openide.util.RequestProcessor$Task.run(RequestProcessor.java:1432) at org.openide.util.RequestProcessor$Processor.run(RequestProcessor.java:2042) WARNING [org.netbeans.core.TimableEventQueue]: too much time in AWT thread org.netbeans.modules.sampler.InternalSampler@77434274
Created attachment 135092 [details] IDE log
If you can reproduce the issue repeatedly, could you please run the self-profiler (alt-shift-y) prior to making the comment ? Save the snapshot & attach here, please. There is a lot of Schema2BeansRuntimeExceptions but not so much they could cause the observed slowdown.
Created attachment 135099 [details] Self sampler I've optimized Garbage Collection, and the slowdown has reduced. But the error is still happening. See attached a self-sampler of the issue.
Created attachment 135108 [details] Better sampler
Was something else running on the machine ? Background build, antivirus software, some disk check, ... ? OS file operations take a lot of time in the snapshot and some are visible even in the log, e.g. WARNING [org.netbeans.modules.masterfs.filebasedfs.utils.FileChangedManager]: 1,950 ms in new File("C:\Users\richter\AppData\Local\NetBeans\Cache\dev\index\s1227\TaskListIndexer\2").exists() The EDT is blocked by JPAProblemFinder; it should run in a separate thread, but that does fix the performance defect only makes the IDE more responsive.
(In reply to comment #5) > Was something else running on the machine ? Background build, antivirus > software, some disk check, ... ? OS file operations take a lot of time in the > snapshot and some are visible even in the log, e.g. I have Microsoft Security Essentials running in background - but I've disabled monitoring of NetBeans, AppData and Project folders to avoid extra hassle. Also, I'm running Tomcat and PostgreSQL - but this "freeze" doesn't happen with DEV 201305222300, only after updade. Maybe I've some corrupted index from previous version? How can I clean only indexes without loosing all my current configuration? > WARNING [org.netbeans.modules.masterfs.filebasedfs.utils.FileChangedManager]: > 1,950 ms in new > File("C:\Users\richter\AppData\Local\NetBeans\Cache\dev\index\s1227\TaskListIndexer\2").exists() Indeed, when NetBeans "freeze", disk activity go high. Also, it stay repeating the "Refreshing indexes" all the time (I can see in the activity bar at down right corner). When I open the "Action Items" window, then it becomes almost unusable. > > > The EDT is blocked by JPAProblemFinder; it should run in a separate thread, but > that does fix the performance defect only makes the IDE more responsive. Does fix or Doesn't fix? I did not understand your last paragraph.
there was no changes in persistence area this week as I remember. it seems quite slow if File.exists take almost 2 seconds, there may be some changes in antivurus of some other background activity. is cache and user directory both in antivirus exclusion list? regarding cache director, see you message.log for 'Cache Directory'. move to separate thread may help with ide response but if there is hdd overload it may stuck at some other point, just the same as if mozilla overload disk, I can't work with netbeans even it's in another thread. But I need to evaluate the issue, as there may be some unnecessary chacks and also do not block edt thread.
also there may be some changes in your project, or in open files or in open tabs.
(In reply to comment #7) > there was no changes in persistence area this week as I remember. > it seems quite slow if File.exists take almost 2 seconds, there may be some > changes in antivurus of some other background activity. is cache and user > directory both in antivirus exclusion list? Cache and user directory are both in exclusion list. > > regarding cache director, see you message.log for 'Cache Directory'. I'll look at it. > > move to separate thread may help with ide response but if there is hdd overload > it may stuck at some other point, just the same as if mozilla overload disk, I > can't work with netbeans even it's in another thread. Seems unlikely. This is a very fast machine (AMD 8 core 3.1Ghz, Sata 3 disks high speed, 8Gb kingston memories, decent ASUS motherboard) - and as I said, started after update to DEV 201305282300. I can move back to 201305222300, but I see no gain for that. > But I need to evaluate the issue, as there may be some unnecessary chacks and > also do not block edt thread. Ok. Let me know if you me to test anything. I've reduced a bit the slowdown with the following customization in my netbeans.conf: netbeans_default_options="--locale en:US -J-Xmx2g -J-client -J-XX:+UseParNewGC -J-XX:ParallelGCThreads=4 -J-XX:MaxGCPauseMillis=500 -J-XX:+DisableExplicitGC -J-XX:+PrintCommandLineFlags -J-XX:+UseConcMarkSweepGC -J-XX:+CMSClassUnloadingEnabled -J-XX:+CMSPermGenSweepingEnabled -J-Dnetbeans.logger.console=true -J-ea -J-Dapple.laf.useScreenMenuBar=true -J-Dapple.awt.graphics.UseQuartz=true -J-Dsun.java2d.noddraw=true -J-Dsun.java2d.dpiaware=true -J-Dsun.zip.disableMemoryMapping=true -J-Dplugin.manager.check.updates=false -J-Dnetbeans.extbrowser.manual_chrome_plugin_install=yes" Originally, I did not customization, and then NetBeans was always getting out of memory while compiling source's folder.
Would this affect my performance: "WARNING [org.netbeans.modules.masterfs.watcher.Watcher]: Cannot add filesystem watch for E:/: java.io.IOException: Unable to open null: 3" All my projects are under drive E:
Do you mean heap increase? sometimes memory lack may cause slowness also. Regarding watcher, I don't know but it may be something related to next item with recent projects ans shouldn't affect performance in this case.
Initial use case have a problem with performance right after <!-- addition but "better sampler" have main problem after "save" action, what is real use case?
Yes, I made tests with DEV 201306032300, and the problem is related to save action. Usually, I type fast, and use keyboard shortcuts for actions (like save), that's why my perception that the problem was with editing speed.
This old bug may not be relevant anymore. If you can still reproduce it in 8.2 development builds please reopen this issue. Thanks for your cooperation, NetBeans IDE 8.2 Release Boss