Re: RPGLE call Java throws MCH74A5

[Scott Klement <rpg400-l@xxxxxxxxxxxxxxxx>]

Justin,

You can only load the JVM once per job.  So when the JVM ends, you won't 
be able to use it again.

I know the error mentions System.exit() (which kills the JVM) but in my 
experience, this usually happens because of a coding error that causes 
Java to terminate itself.  (For example, if you pass a *NULL object, 
Java will terminate itself, and subsequent messages will refer to 
System.exit)

You should examine what's happening during the first call, see if 
something is happening that is causing an error, and try to fix that code.

I guess it's possible that the 3rd party app is calling System.exit() 
directly -- but that would surprise me greatly, as I've not seen a Java 
app do that, and I've interfaced with quite a few Java apps!  If the app 
really is calling System.exit(), then RUNJVA might be your best bet. 
Since the JVM is ended with each call, you won't be able to improve upon 
performance.  But, imho, this is unlikely to be the case.

If performance is important, you might find it performs better if Java 
and this application are loaded into a separate job, such as a Java 
application server, and then connect and communicate with that job 
through sockets, pipes, data queues, or something like that.  Obviously 
this adds a new level of complexity, but will give you the best performance.

Calling Java from RPG (which uses JNI) may be better performance than 
calling RUNJVA repeatedly, though.

-SK


On 11/27/2015 10:16 AM, Justin Taylor wrote:
> I have an RPGLE program that calls a 3rd party Java app.  Right now it does a handful of RUNJVA using QCMDEXC.  I'm trying to call the Java directly from RPGLE because I think I'll get better performance on repeat calls.  Running interactively, the first Java call seems to work, but I get this MCH74A5 error:
> Message . . . . :   The Java Virtual Machine has ended.
> Cause . . . . . :   Java Virtual Machine 1 has ended because of reason 1.  The
>    reason codes are defined as follows:
>      01- A Java program called the java.lang.System.exit method with a zero
>    status code.
>
>  From what I can find, a 0 code means success, so I'm not sure why it's killing my JVM.