This is my first attempt at using the Memory Analyzer tool and I'm having
some problems. I'm trying to analyze a dump from IBM WebSphere v6.0 and
I'm getting results I don't expect. No size information is loaded with
the dump. The Overview Details state:
"Size: 0 B Classes: 15k Objects: 16k Class Loader: 894"
Here's the output of "java -version" from WAS's JVM:
java version "1.4.2"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2)
Classic VM (build 1.4.2, J2RE 1.4.2 IBM AIX build ca142-20060421 (SR5)
(JIT enabled: jitc))
I'm using the latest versions of Memory Analyzer and IBM DTFJ.
Am I doing something wrong when generating the dump? Do I have a version
incompatibility? Any help would be greatly appreciated.
Are you processing a Jextracted core file (.sdff) or a javacore file? A
Javacore file is a readable text file with some limited information about
the JVM - classes, class loaders, threads etc. There isn't any useful
object information in a javacore file, so the sizes will be zero and
nearly every object will be missing. That is why the help says:
The DTFJ adapter can process Javadumps (javacore.txt files) generated
using the -Xdump:java command-line option. This is not very useful within
MAT because the only information in the Javadump is a list of classes and
class loaders. There is no object or size information.