Andy Wu said the following on 10/03/2012 11:53 AM:
David,
Sorry I misread your email first time round. Ignore what I said
previously.
Let me give it try here and see what I can spot.
I ran the following test case on Linux (Java6 3GB heap, 640m perm gen)
with OpenBD nightly:
<cfcomponent>
<cffunction name = 'test' access='remote' returnformat='json'>
<cfset var ii = 0 />
<cfset var i = 0 />
<cfloop from="1" to="500000" index="ii">
<cfloop from="1" to="5" index="i">
<cfthread action="run" name="thread#ii#-#i#"
outer="#ii#" inner="#i#">
<cfset console("running thread number #outer# -
#inner#") />
</cfthread>
</cfloop>
<!--- We don't have to join individual threads, but all 5
at once. Also no need to terminate them manually --->
<cfthread action="join"
name="thread#ii#-1,thread#ii#-2,thread#ii#-3,thread#ii#-4,thread#ii#-5" />
</cfloop>
<cfreturn "done ;)" />
</cffunction>
</cfcomponent>
Getting:
Exception in thread "Timer-12" java.lang.OutOfMemoryError: GC overhead
limit exceeded
at org.aw20.util.SystemClock.setTime(SystemClock.java:82)
at org.aw20.util.SystemClock.access$000(SystemClock.java:35)
at org.aw20.util.SystemClock$1.run(SystemClock.java:67)
at java.util.TimerThread.mainLoop(Timer.java:512)
at java.util.TimerThread.run(Timer.java:462)
Running a VisualVM heap dump. I see that perm gen space is fine (about
250mb). However I'm seeing a stack up of javolution.util.FastMap$Entry
in the instances (2,208,894 instances) which keeps going up, but never
recovered.
--
Peter J. Farrell
OpenBD Steering Committee / Mach-II Lead Developer
http://blog.maestropublishing.com
Identi.ca / Twitter: @maestrofjp
Please do not send me Microsoft Office/Apple iWork documents. Send OpenDocument
instead! http://fsf.org/campaigns/opendocument/
--
online documentation: http://openbd.org/manual/
http://groups.google.com/group/openbd?hl=en
