When the problem starts (IO activity, mem rising)
threads (of the JD java
process) are created and deleted in an endless loop, in a rate of one every two seconds cycle aprox -- it could explain the strange IO behav.
The
thread
Code:
msvcr71.dll!endthreadex+0x31
and its
stack
Code:
ntoskrnl.exe!ZwAssignProcessToJobObject+0x15
ntoskrnl.exe!KeQueryRuntimeThread+0x5e8
ntoskrnl.exe!CcPurgeCacheSection+0x240
ntoskrnl.exe!NtQueryInformationToken+0xace
ntoskrnl.exe!ZwSetSystemInformation+0x23
ntdll.dll!KiFastSystemCallRet
kernel32.dll!WaitForSingleObject+0x12
jvm.dll!JVM_FindSignal+0x50b25
jvm.dll!JVM_FindSignal+0x92595
jvm.dll!JVM_FindSignal+0x929a7
jvm.dll!JVM_MonitorWait+0xaf
jvm.dll!AsyncGetCallTrace+0x3facc
jvm.dll!JVM_FindSignal+0x53111
jvm.dll!AsyncGetCallTrace+0x3fc97
jvm.dll!AsyncGetCallTrace+0x3fd0d
jvm.dll!jio_fprintf+0x140
jvm.dll!JVM_FindSignal+0xaf354
jvm.dll!JVM_FindSignal+0x52d8c
msvcr71.dll!endthreadex+0xa0
kernel32.dll!GetModuleFileNameA+0x1ba
This happens soon after the JD start (some minutes), particularly when its window is visible (not minimized).
Perhaps the memory leak is the sum of unrecoverable bytes allocated in this cyclic, apparently misguided, behavior.
Edited: As said before, the amount of allocated memory increases more rapidly when the window is not minimized.