Mailing-List: contact tomcat-user-help@jakarta.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Tomcat Users List" <tomcat-user@jakarta.apache.org>
From: "Rob Wichterman" <rwichterman@nuventive.com>
To: <tomcat-user@jakarta.apache.org>
Subject: Odd Memory Problem
Date: Fri, 2 Apr 2004 11:17:07 -0500
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_0016_01C418A4.09569970"
Thread-Index: AcQYzfG8vIq+rL+WT+q5qKN5hVTYWA==
Message-Id: <20040402161257.VGDQ15204.imta01a2.registeredsite.com@rwichterman>

------=_NextPart_000_0016_01C418A4.09569970
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit

I have been battling a memory problem for a few months now with no luck.  I
have used profilers and everything else you could imagine.  The problem is
we have never been able to duplicate it on the testing servers it only
happens on production which does not see much load at all (100 users a day).
I noticed something very odd yesterday though.  I am using the
"-XX:+UseConcMarkSweepGC and -Xms512m". I was monitoring the garbage
collection activity very closely and I was receiving good performance for 3
days.  I was seeing one garbage collection about every 5 minutes in the
Catalina.out.  Then out of nowhere the garbage collection starting going
crazy, about one garbage collection a second, I checked the cpu usage and it
was stuck around 60% used.  There were only 2-4 people logged in at a time.
The crazy garbage collection went on for a couple of hours before it
eventually crapped out with "Out of Memory" errors.  It almost seems like a
section of the app caused this sudden outburst but there is no way our QA
department hasn't went over the same section.  I was hoping some one seen or
heard of this before.  Below is my environment.

 
Thanks,

 
Apache2.0.47 using mod_jk2

2 * tomcat4.1.24

J2sdk1.4.2_04


------=_NextPart_000_0016_01C418A4.09569970--