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 205282 - Improve duplicate detection for slownesses
Summary: Improve duplicate detection for slownesses
Status: NEW
Alias: None
Product: ide
Classification: Unclassified
Component: Logger Server (show other bugs)
Version: 7.1
Hardware: PC Mac OS X
: P1 normal (vote)
Assignee: Tomas Danek
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2011-11-18 10:31 UTC by Tomas Danek
Modified: 2013-07-11 10:01 UTC (History)
2 users (show)

See Also:
Issue Type: ENHANCEMENT
Exception Reporter:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Tomas Danek 2011-11-18 10:31:40 UTC
see evaluation in https://netbeans.org/bugzilla/show_bug.cgi?id=194908
Comment 1 Jesse Glick 2011-11-18 16:28:03 UTC
Unfortunately it is a hard problem. In bug #194908, the code which is slow is in fact similar across all the reports. This code is just inherently (potentially) expensive, so the onus is on callers to avoid blocking on it. It probably takes a human to know that, though.
Comment 2 Jesse Glick 2011-11-18 16:33:27 UTC
I think the real problem may be in the policy of trying to group reports and automatically file bugs for them. The bugs are extremely time-consuming to analyze, and the result is often that the reports are misgrouped, or that it is simply hard to know what the real problem was (e.g. IDE was swapped out of RAM and everything was grinding to a halt).

I would suggest just collecting reports, then having a small team pick out randomly selected reports (biasing those with longer delays), determining if the IDE is really at fault (and also whether there is any chance of reproducing the problem), and manually filing a bug in the responsible component with a link to the report and a summary of the analysis.
Comment 3 Marian Mirilovic 2013-07-11 10:01:58 UTC
(In reply to comment #2)
> I think the real problem may be in the policy of trying to group reports and
> automatically file bugs for them. The bugs are extremely time-consuming to
> analyze, and the result is often that the reports are misgrouped, or that it is
> simply hard to know what the real problem was (e.g. IDE was swapped out of RAM
> and everything was grinding to a halt).

I can't agree more !