On 10/04/2012 23:21, Doron Tsur wrote: > Another conclusion to share here is that when running on windows 2003 > enterprise R2 (the one we tested, but could be other windows OS) killing > tomcat will not return the non page pool back to the OS. Even on systems > that it isn't leaking. Usually there is no need for so much memory > including our system, but this should be checked.
You should use a tool like VisualVM or JConsole to inspect the running JVM and try to identify if one of the memory areas of the JVM increases in parallel with the leak you're seeing from the OS view. p > PID- thanks for the tip. > > > On Tue, Apr 10, 2012 at 5:31 PM, Pid <p...@pidster.com> wrote: > >> On 10/04/2012 13:11, Doron Tsur wrote: >>> Switching to Blocking IO from Non Blocking IO seems to have solved the >>> issue. I don't know why, does anybody have ideas? >> >> Try the latest version of Tomcat. There were patches to the Connector >> code in versions subsequent to 7.0.23. >> >> >> p >> >> >> -- >> >> [key:62590808] >> >> > -- [key:62590808]
signature.asc
Description: OpenPGP digital signature
