JR50653: JAVA INTEGRATION STEP FAILS WHEN HUGE JAR FILES ARE INVOLVED

Version 8.5 Refresh Pack 6 for the IBM Business Process Manager products
Version 8.5.0 Fix Pack 2 for the IBM Business Process Manager products

APAR status

• Closed as program error.

Error description

• You can model a Java integration step in a service flow to run
  arbitrary Java methods. However, the Java classes that are
  required to run these methods must be attached to the process
  application or toolkit as managed assets (server files).
  There is a special class loader that manages loading classes
  from managed assets.  Sometimes, the classes are swapped out of
  the cache and, as a result,  the classloader of the managed
  assets returns null instead of the resource data. The
  exception that you see depends on the Java method that is being
  called. The exception might indicate that a class-loading issue
  has occurred. For example, Java classes cannot be loaded
  correctly or property files are not loaded.
      This could also result in a temporary performance problem
  while the classes in the jar are being loaded.
  It may result in a WSVR0605W message like this
  [11/21/14 12:00:14:584 EST] 0000001f ThreadMonitor W
  WSVR0605W:
  Thread "WebContainer : 132" (00004752) has been active for
  760075 milliseconds and may be hung.  There is/are 199
  thread(s) in total in  the server that may be hung.
      at
  com.lombardisoftware.server.core.ManagedAssetClassLoader.loadCla
  ss(ManagedAssetClassLoader.java)
      at java.lang.ClassLoader.loadClass(ClassLoader.java:665)
      at org.mozilla.javascript.Kit.classOrNull(Kit.java:92)
      at
  org.mozilla.javascript.Kit.testIfCanLoadRhinoClasses(Kit.java:12
  1)
      at
  org.mozilla.javascript.Context.getApplicationClassLoader(Context
  .java:2273)
      at
  org.mozilla.javascript.NativeJavaTopPackage.init(NativeJavaTopPa
  ckage.java:106)
      at sun.reflect.GeneratedMethodAccessor306.invoke(Unknown
  Source)
      at
  sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethod
  AccessorImpl.java:37)
      at java.lang.reflect.Method.invoke(Method.java:611)
      at
  org.mozilla.javascript.ScriptableObject.buildClassCtor(Scriptabl
  eObject.java:1030)
      at
  org.mozilla.javascript.LazilyLoadedCtor.buildValue(LazilyLoadedC
  tor.java:110)
      at
  org.mozilla.javascript.LazilyLoadedCtor.init(LazilyLoadedCtor.ja
  va:89)
      at
  org.mozilla.javascript.ScriptableObject.getImpl(ScriptableObject
  .java:2032)
      at
  org.mozilla.javascript.ScriptableObject.get(ScriptableObject.jav
  a:287)
      at
  org.mozilla.javascript.IdScriptableObject.get(IdScriptableObject
  .java:387)
      at
  org.mozilla.javascript.ScriptableObject.getProperty(ScriptableOb
  ject.java:1617)
      at
  org.mozilla.javascript.ScriptRuntime.topScopeName(ScriptRuntime.
  java:1788)
      at
  org.mozilla.javascript.ScriptRuntime.name(ScriptRuntime.java:169
  0)
      at
  org.mozilla.javascript.Interpreter.interpretLoop(Interpreter.jav
  a:3410)
      at
  org.mozilla.javascript.Interpreter.interpret(Interpreter.java:24
  87)
      at
  org.mozilla.javascript.InterpretedFunction.call(InterpretedFunct
  ion.java:164)
      at
  org.mozilla.javascript.ContextFactory.doTopCall(ContextFactory.j
  ava:398)
      at
  com.lombardisoftware.core.script.JavaScriptContextFactory.doTopC
  all(JavaScriptContextFactory.java:156)
      at
  org.mozilla.javascript.ScriptRuntime.doTopCall(ScriptRuntime.jav
  a:3070)
      at
  org.mozilla.javascript.InterpretedFunction.exec(InterpretedFunct
  ion.java:175)
      at
  org.mozilla.javascript.Context.evaluateString(Context.java:1104)
      at
  com.lombardisoftware.core.script.js.PreparedScope.createStandard
  ObjectScope(PreparedScope.java:656)
  

Local fix

• You can prevent this issue from recurring by increasing the size
  of the managed asset class loader's resource map by adding the
  following lines to the 100Custom.xml file:
  
  <server> <|-- add if not already present-->
    <classloader-resource-map-size>size
    </classloader-resource-map-size>
  </server><|-- add if not already present-->
  
  The actual size depends on the total number of resources that
  are in your managed asset jar files.
  

Problem summary

• The managed asset class loader maintains a cache of resources
  that are already loaded.
  If a managed asset jar file contains more resources than the
  maximum cache size can accept, a resource might not be loaded
  correctly and ?null? might be returned.
  

Problem conclusion

• A fix is available that fixes this class loading issue.
  
  On Fix Central (http://www.ibm.com/support/fixcentral), search
  for JR50653:
  
  1. Select IBM Business Process Manager with your edition from
    the product selector, the installed version to the fix pack
    level, and your platform, and then click Continue.
  
  2. Select APAR or SPR, enter JR50653 , and click Continue.
  
  When you download fix packages, ensure that you also download
  the readme file for each fix. Review each readme file for
  additional installation instructions and information about the
  fix.
  

Temporary fix

• Not applicable
  

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

• APAR is sysrouted TO one or more of the following:

Fix information

• Fixed component name

  BPM STANDARD

• Fixed component ID

  5725C9500

Applicable component levels

• R850 PSY

     UP

• R855 PSY

     UP