> On Thu, Mar 04, 2010 at 11:31:24AM -0500, Carson, Brad wrote: >> >> I was hoping that someone else had seen this type of issue before >> and might provide some more insight. Our application folks keep >> wanting to blame z/VM and the z10 hardware instead of looking at >> their code.
Brad, As a gross generalization, poor performance of an application using an application server architecture is seldom the fault of the application server. It's *almost* always something the application is doing that causes the application server to misbehave (there are exceptions if the underlying hardware is inadequate). Similar issues are present with Websphere and Jboss -- the app server can only do what it's told to do, and there are a lot of programs that tell it to do things that don't make sense in a shared-resource environment. > You really want to point some decent Java monitoring > tools at your JVM to help you understand what's going on here. This is the answer you're looking for. Keep in mind that CPU numbers are just going to be relative comparisons (few of the Java monitoring tools have any idea of correlation of resources in hypervisors, although a few have rudimentary support for Xen and Vmware). The time spent in particular routines will tell you more about an application performance problem than just comparing CPUs. ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
