Some fullGC do not appear as such with CMS

[acailley]

Hi !

with CMS line
742285,854: [Full GC 742285,856: [CMS742292,433: [CMS-concurrent-mark: 9,640/12,329 secs] [Times: user=114,49 sys=1,09, real=12,33 secs]
are not counted as a FullGC.

Then we only have a blank in the chart :

[acailley]

The complete file GC_IS_PRJ55310_150511_202134.7z is available at
http://www.filedropper.com/gcisprj55310150511202134

[chewiebug]

Thanks for raising the issue and providing the log file. Usually gcviewer has no problems parsing a full gc mixed with a concurrent event. I'll have to check with the log file.

I am on vacation until end of May, you'll have to be patient.

Best regards,
Jörg

-------- Original Message --------
From: acailley notifications@github.com
Sent: May 22, 2015 4:41:44 PM CEST
To: chewiebug/GCViewer GCViewer@noreply.github.com
Subject: Re: [GCViewer] Some fullGC do not appear as such with CMS (#135)

The complete file GC_IS_PRJ55310_150511_202134.7z is available at
http://www.filedropper.com/gcisprj55310150511202134

---

Reply to this email directly or view it on GitHub:
#135 (comment)

[chewiebug]

Sample log file can now be parsed without warnings, full gcs are recognised. The latest SNAPSHOT version of GCViewer contains the fix.

Thanks for raising this issue!

Best regards,
Jörg

[acailley]

Yes, looks like I did not have the last build/sources, a stupid mistake.
With the last build, I have no error and the fullGC are correctly displayed.

Thanks a lot for your work, sorry again for loosing your time.

[chewiebug]

Ah, well, no! I have indeed found a bug based on your issue description and published a fixed version. I just suspect, that you used an older version of the code for your analysis. You didn't loose my time, you helped me improve GCViewer! Sorry for the confusion, I created! Best regards, Jörg