This forum is closed to new posts and
responses. Individual names altered for privacy purposes. The information contained in this website is provided for informational purposes only and should not be construed as a forum for customer support requests. Any customer support requests should be directed to the official HCL customer support channels below:
The Java maximum heap size should be set to the appropriate size for the application, not the machine.
If you machine has 4 times as much memory as another machine that doesn't mean Notes needs or should use that extra memory.
When the maximum heap is increased, that encourages the Java VM to use that memory. In other words, the VM doesn't work very hard to not grow the heap. That's why it's important to pick a value that works for the application.
It's unlikely that increasing the maximum Java heap will help most folks. One case where it could help is if your heap is mostly full (within 10% of the specified -Xmx which is 256MB) and stays nearly full. As you heap size approaches the maximum, the garbage collector will spend more and more time trying to collect/compact the large heap.
In my opinion, if your Java heap is constantly bumping up against the 256MB limit there are bigger problems, and simply increasing the maximum isn't the answer. If you're running into situations where your heap grows that large I consider it a bug and we need to get more information and find/fix the defect.
In general, the more knobs you start tweaking in the JVM, the more you prevent the JVM from doing what it thinks is best because you're telling it that you know better than it does. Sometimes that's the right thing to do, other times it isn't. The best way to answer these questions is with hard data in a controlled environment.
Back in Notes 8 we did not set a JVM maximum heap size which means that the VM used half of physical memory. Which means if you have 2 GB of RAM it was like you'd set the maximum heap to 1 GB.
No one praised our performance in Notes 8 with this setting and in fact we were criticized for our large memory footprint and slow startup times.
--
Chris Grindstaff
Technical Lead - Notes Performance
Performance Analyst - Java Client Performance
Feedback response number CGRF8A7GH6 created by ~Elizabeth Frovelusonjip on 10/13/2010
A popular question
Return to top
. . A popular question (~Joan Dwohipiad... 13.Oct.10) 
