[07:55:39 EDT(-0400)] * jayshao (n=jayshao@jayshao.rutgers.edu) has joined ##uportal
[08:57:10 EDT(-0400)] * esm (n=esm@asdf.dkc.jhu.edu) has joined ##uportal
[09:58:45 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[09:58:53 EDT(-0400)] <EricDalquist> good morning
[10:39:55 EDT(-0400)] <EricDalquist> Anyone else awake?
[10:43:46 EDT(-0400)] * pberry (n=pberry@waldorf.CSUChico.EDU) has joined ##uportal
[10:44:15 EDT(-0400)] <pberry> okay, who narc'd on me yesterday? I got busted for talking about CAS (it was really more about maven)
[10:44:28 EDT(-0400)] <pberry> and by "busted" I mean "Scott B. tried to help me."
[10:44:32 EDT(-0400)] <EricDalquist> lol
[10:50:16 EDT(-0400)] <pberry> http://blogs.sun.com/watt/resource/jvm-options-list.html
[10:50:27 EDT(-0400)] <pberry> that's a nice list
[11:23:37 EDT(-0400)] * jayshao_ (n=jayshao@wg-206-hill012.rutgers.edu) has joined ##uportal
[11:31:04 EDT(-0400)] <EricDalquist> that it is
[11:31:09 EDT(-0400)] <EricDalquist> adding it to the bookmarks
[11:31:23 EDT(-0400)] <EricDalquist> I need to go and do more JVM/GC research on 1.6 one of these days
[11:34:53 EDT(-0400)] <pberry> I'm starting to see very abnormal memory usage
[11:34:57 EDT(-0400)] <pberry> it's staying very low
[11:35:11 EDT(-0400)] <pberry> and not going into the spike/drop cycle like I expect it to
[11:35:28 EDT(-0400)] <EricDalquist> what is the GC doing?
[11:35:46 EDT(-0400)] <pberry> well, I hate to admit it, but I'm not logging GC
[11:35:54 EDT(-0400)] <pberry> here is the 26th http://seaborn.csuchico.edu/stats/portal/?Machine=All&Type=Hours&Field=usedmem&Month=8&Day=26&Year=2007&Hour=0&Finish=23&Combined=selected&.submit=Submit&.cgifields=Combined&.cgifields=Machine
[11:36:08 EDT(-0400)] <pberry> here was yesterday http://seaborn.csuchico.edu/stats/portal/?Machine=All&Type=Hours&Field=usedmem&Month=8&Day=29&Year=2007&Hour=0&Finish=23&Combined=selected&.submit=Submit&.cgifields=Combined&.cgifields=Machine
[11:36:39 EDT(-0400)] <EricDalquist> the second one is more what we see
[11:36:48 EDT(-0400)] <pberry> really? ah, that's good!
[11:36:57 EDT(-0400)] <EricDalquist> using the parallel new collector and CMS tenured collector
[11:36:59 EDT(-0400)] <EricDalquist> well
[11:37:08 EDT(-0400)] <EricDalquist> it depends what GC approach the JVM is taking
[11:37:14 EDT(-0400)] <EricDalquist> we don't ever see full GCs
[11:37:34 EDT(-0400)] <EricDalquist> the ParNew collector runs very frequently (every minute or so) on our eden space
[11:37:48 EDT(-0400)] <pberry> we keep it pretty simple
[11:37:49 EDT(-0400)] <pberry> CATALINA_OPTS="-server -Xms1024m -Xmx1536m -Dsun.net.inetaddr.ttl=0"
[11:38:11 EDT(-0400)] <EricDalquist> the ConcurrentMarkSweep GC runs constantly in a background thread pruning tenured space
[11:38:19 EDT(-0400)] <EricDalquist> you need to enable JMX
[11:38:34 EDT(-0400)] <EricDalquist> then you can connect using JConsole and see which collectors are being used
[11:38:39 EDT(-0400)] <EricDalquist> and the time spent doing collections
[11:38:49 EDT(-0400)] <EricDalquist> and per-heap space usage
[11:39:47 EDT(-0400)] <pberry> oh. right. /realizes I'm out of my element
[11:39:56 EDT(-0400)] <EricDalquist> 
[11:40:29 EDT(-0400)] <EricDalquist> oi ... the spider web of wiki pages makes my head hurt at times
General
Content
Integrations