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.

Bug 176476 - 11s in HudsonLogger$Helper$3.run() or in ServiceLoader.load(...)
Summary: 11s in HudsonLogger$Helper$3.run() or in ServiceLoader.load(...)
Status: RESOLVED WORKSFORME
Alias: None
Product: connecteddeveloper
Classification: Unclassified
Component: Hudson (show other bugs)
Version: 6.x
Hardware: All All
: P4 normal (vote)
Assignee: Jaroslav Havlin
URL:
Keywords: PERFORMANCE
Depends on:
Blocks:
 
Reported: 2009-11-10 22:41 UTC by Jaroslav Tulach
Modified: 2012-11-07 15:43 UTC (History)
2 users (show)

See Also:
Issue Type: DEFECT
Exception Reporter: 161782


Attachments
nps snapshot (256.00 KB, application/nps)
2009-11-10 22:41 UTC, Jaroslav Tulach
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Jaroslav Tulach 2009-11-10 22:41:09 UTC
This issue was originally marked as duplicate of issue 169801, that is already resolved. This issue is still valid, so this seems to be another issue, but it might be related.

Build: NetBeans IDE Dev (Build 091103)
VM: OpenJDK Client VM, 14.0-b16, OpenJDK Runtime Environment, 1.6.0_0-b16
OS: Linux, 2.6.31.5-server-1mnb, i386

User Comments:
jtulach: Trying to scroll in just opened output window. The window shows failure from a Hudson server


Maximum slowness yet reported was 11985 ms, average is 11985
Comment 1 Jaroslav Tulach 2009-11-10 22:41:14 UTC
Created attachment 90761 [details]
nps snapshot
Comment 2 Jaroslav Tulach 2009-12-01 00:50:21 UTC
Jesse, Nejedláku, can you please look at this snapshot. Originally I thought that the problem is in FoD, but at the end FoD seems an innocent victim here.

#1 - you could change the Hudson$Logger to run outside of AWT. At least I think that Line$Show does not require to be in AWT.

#2 - there is >6s wasted while loading 
META-INF/services/com.sun.org.apache.xml.internal.dtm.DTMManager
the snapshot shows 22 ZIP files being open. Given that this is just sampling it might even be true that the classloading in platform was opening all the JARs. Seems to me that the classloading cache failed in this case.
Comment 3 Petr Cyhelsky 2012-11-07 15:43:07 UTC
latest report from 7.1 seems to work now, reopen if encountered again