[jboss-user] [Performance Tuning] - Variation in Max Heap size

Rohit Macherla do-not-reply at jboss.com
Tue Jul 27 02:03:04 EDT 2010

Rohit Macherla [http://community.jboss.org/people/rohit.macherla] replied to the discussion

"Variation in Max Heap size"

To view the discussion, visit: http://community.jboss.org/message/554552#554552

Hi Ruchir,

Thanks for the response. Please find my response for the questions :

> 01> What type of platform you are using
Rohit: We are using HP Unix B11.23_LR
> 02> What is the total RAM of the Machine
>  03> how many CPUs you have on the box and what type of CPUs you have
Rohit: 4 core CPU, 24GB RAM. Planning to upgrade to 8 core 36GB soon, owing to planned increase in the load.

> 04> please avoid using "-server" as a input parameter to the JVM, JDK1.5 onwards dose it automatically for you, it checks the os to get this value
Rohit: Okay. Will test it.

> 05> Based on the details which you have provided i think this will work
>  -Xmx2048m -Xms2048m -XX:MaxNewSize=256m -XX:NewSize=256m -XX:MaxPermSize=128m -XX:SurvivorRatio=8 -XX:TargetSurvivorRatio=90 -XX:MaxTenuringThreshold=0 -XX:CMSInitiatingOccupancyFraction=60 -XX:LargePageSizeInBytes=5m -XX:ParallelGCThreads=20   -XX:+StringCache -XX:+UseParNewGC -XX:+UseConcMarkSweepGC -XX:ThreadStackSize=1024 -Dsun.rmi.dgc.client.gcInterval=3600000 -Dsun.rmi.dgc.server.gcInterval=3600000
>  06> Try to go through this articel " http://community.jboss.org/docs/DOC-15623?uniqueTitle=false http://community.jboss.org/docs/DOC-15623?uniqueTitle=false" we have tried to explain it.
Rohit: I am half way through the document. We'll make the changes suggested in  amock environment for testing. However, we are more intent on why the heap size is changing.

As a general follow up to the chain, after some hiccups with our support with vendor, we were asked to submit a more recent version of the event and we are in the process of finding a relationship between the max heap variation and CPU utilization. Recently discovered a better way to find the heap statistics using twiddle. I'll keep the results posted after our testing.

Hi Peter,
As far as I know, having remote objects garbage collected will result in a full GC and if an application does not have much involvement for remote objects, the interval of collection can be set to a higher value. I have no idea how to measure whether the heap I have uses remote objects. Currently, we use webservices and all our apps reside on the same JVM. How the external systems access our webservices, whether it creates remote objects or not is still a black area for me. Since I explicitly do not use RMI, I have enabled a one hour DGC.
I will greatly appreciate if you can provide some inputs on the same.

Rohit M.


Reply to this message by going to Community

Start a new discussion in Performance Tuning at Community

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.jboss.org/pipermail/jboss-user/attachments/20100727/187aabd9/attachment.html 

More information about the jboss-user mailing list