1
It would be really helpful for me if someone can answer my question.
How does Kernel allocate COMMITTED VIRTUAL MEMORY to a process (jvm process)
what criteria is used to determine how much COMMITTED VIRTUAL MEMORY is required by a process.
My current Application server : websphere Ram:3.5Gb Heap: 2Gb
When websphere starts up my, COMMITTED VIRTUAL MEMORY is 5.1 Gb .
When websphere is taking a load, COMMITTED VIRTUAL MEMORY is 5.5 Gb, at this point Kernel is killing the Jboss EAP PID.
As a temporary fix i increased my RAM to avoid termination of EAP process id by kernel.
But i would like to find out the criteria Kernel is using to allocate COMMITTED VIRTUAL MEMORY.
There's quite a bit of documentation on Virtual Memory allocation in the Linux kernel. Have you looked this up or read it yet? Are there parts that aren't making sense to you? https://www.kernel.org/doc/Documentation/sysctl/vm.txt https://www.linuxjournal.com/article/1133 https://www.win.tue.nl/~aeb/linux/lk/lk-9.html
– music2myear – 2019-06-20T15:24:04.783