| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
| Running a Java Class file |
« View previous topic :: View next topic » |
| Author |
Message
|
| Vladimir |
 Posted: Tue Nov 19, 2002 11:38 am Post subject: |
 |
|
Acolyte
Joined: 14 Nov 2002
Posts: 73
Location: USA, CA, Bay Area
|
It seems like I finally got JHPB working.
IBM was pushing me wrong directions and JHPB was never started on my box, but they were saying it was running... 
The problem was in several things:
- use IBM JDK 1.3.1 or Sun SDK at least 1.3.1_04. IBM said that Sun JRE is enough, but it is not true. The truth is that they want to use JRE from SDK and it is different from JRE that is shipped as a standalone package.
- just proper configure PATH, CLASSPATH and JAVA_HOME (I am not sure about the minimum configuration you should do, that's why I am not putting my suggestions here. I have a lot of stuff added to this variables, I am sure it is an overkill)
But it works!!!  |
|
| Back to top |
|
 |
| muralihegde |
| Posted: Tue Nov 19, 2002 10:24 pm Post subject: |
|
|
Centurion
Joined: 30 Apr 2002
Posts: 108
|
Hi,
I have a java program as part of the activity implementation in the process model. On NT WF I filled in the tab with the path of my java class file to be invoked. I have my class file in the ..progarm file\mqseries work flow\bin folder so that it is available in the path
The java pgm is invoked without any problem on my NT.
But in case of S/390, we have to set up the parameters on the OS/390 tab on the program property which needs to be entered for customizing the PES Directory. (Customization & Admin manual page no. 101 , 102 109).
The parameters are like Service, Invocation type, service type etc..
The customization manual gives an example for CICS, IMS etc and for user defined programs direct us to fill in the appropriate parameters. I am finding it difficult to compare these parameters.
Will it be possible for you to advise me what these parameters should be if I have to invoke a simple java pgm on the USS or if I decide to invoke a JCL which in turn calls the java program.
Thanks...
-Murali |
|
| Back to top |
|
|
| Hari |
| Posted: Thu Nov 21, 2002 5:55 am Post subject: Approx. similar problem as Mike |
|
|
Centurion
Joined: 21 Nov 2002
Posts: 117
Location: USA
|
Hi all,
I am facing a approximately same problem as Mike.
Well my problem is that i have simple Java application(using Frame) which i need to call.
Well i am not getting any errors, but when i start off the activity associated with the program then [b]the "Frame(Java application)" does not appear[/b], but still the process moves on.
As a standalone the application works really fine so i dont think there's any problem with the application.
Please advice,
Thx in advance,
Hari |
|
| Back to top |
|
|
| jmac |
| Posted: Thu Nov 21, 2002 6:06 am Post subject: |
|
|
Jedi Knight
Joined: 27 Jun 2001
Posts: 3081
Location: EmeriCon, LLC
|
Hari:
We need a little more information to understand what you are doing.
If the process is moving on, then the APP is running. MQWF would not advance the instance to the next activity, unless the current activity ended and satisfied it's exit condition.
_________________
John McDonald
RETIRED |
|
| Back to top |
|
|
| Vladimir |
| Posted: Thu Nov 21, 2002 4:10 pm Post subject: |
|
|
Acolyte
Joined: 14 Nov 2002
Posts: 73
Location: USA, CA, Bay Area
|
John, Hari:
If the process is moving on - it doesn't really mean that program was launched successfully. I do not have any experience with WF under 390 (that's why I cannot give any advice here), but under NT when you use Java High Performance Bridge, in case JHPB cannot find the JVM to start - it will return result code = 0 and move on the process. I know, it sounds not right, but that's the way JHPB works now. I assume, that other bridges (or launch facilities) can behave the same way. I hope it is not true in your case, but I just want to warn you that such thing can happen under WF. |
|
| Back to top |
|
|
| jmac |
| Posted: Thu Nov 21, 2002 4:48 pm Post subject: |
|
|
Jedi Knight
Joined: 27 Jun 2001
Posts: 3081
Location: EmeriCon, LLC
|
Vladimir:
If what you say about the JHPB is true, then in my opinion there is a bug. If the activity does not return successfully, the workflow engine should not move the process on. What I am hearing you say, is that the activity is never even found in the first place, but the Workflow engine says it ran successfullly! This would be a major problem and a reason to never use the JHPB.
_________________
John McDonald
RETIRED |
|
| Back to top |
|
|
| Vladimir |
| Posted: Thu Nov 21, 2002 5:56 pm Post subject: |
|
|
Acolyte
Joined: 14 Nov 2002
Posts: 73
Location: USA, CA, Bay Area
|
John,
It sounds like a bug for me too. But it is in real that was making our recent conversations with IBM very complex. They were saying that JHPB is running, because of the fact that it was returning successful result code, but I saw in the trace that it cannot fin a JVM and obviously not executing my program.
That's why I will probably open another PMR with IBM on this issue if I set up my test environment according to their recommendation and I will have this issue still. But I am getting too busy now and not sure when it happens...  |
|
| Back to top |
|
|
| Vladimir |
| Posted: Thu Nov 21, 2002 10:51 pm Post subject: |
|
|
Acolyte
Joined: 14 Nov 2002
Posts: 73
Location: USA, CA, Bay Area
|
Here is an update on our tests with JHPB.
Good news:
We were wrong.
JHPB returns _RC=-1 if VM cannot be found.
It was a problem in the test model.
Bad news:
It makes IBM support look even worse, because they were saying (and are saying) that "based on the trace logs JHPB is running fine". 
Conclusion:
Trust noone.  |
|
| Back to top |
|
|
| vennela |
| Posted: Thu Nov 21, 2002 11:29 pm Post subject: |
|
|
Jedi Knight
Joined: 11 Aug 2002
Posts: 4055
Location: Hyderabad, India
|
Vladimir:
| Quote: |
| They were saying that JHPB is running, because of the fact that it was returning successful result code, but I saw in the trace that it cannot fin a JVM and obviously not executing my program. |
| Quote: |
| JHPB returns _RC=-1 if VM cannot be found. |
These are two contradicting statements you are making here. When you provide information saying that something is returning an error code 0 (they don't care what it is ..) then they would definitely believe that the product side of it works just fine and they would grow more suspicious about your app. It's not fair to blame IBM's support when you have not done your home work properly.
We should learn from the mistakes and use the IBM's support after mqseries.net  doesn't provide us the solution (also beacuse the response here is lightening).
---
Venny |
|
| Back to top |
|
|
| Vladimir |
| Posted: Fri Nov 22, 2002 12:10 am Post subject: |
|
|
Acolyte
Joined: 14 Nov 2002
Posts: 73
Location: USA, CA, Bay Area
|
Venny,
You are cheating
You didn't quote my statement "We were wrong".
It means some of our results were wrong.
Yes, we got an impression that no error code was returned, but as I said "It was a problem in the test model." and we were not checking it properly. I found and corrected this mistake and saw that _RC is -1 in that case.
Yes, I know that it sounds like I blame IBM in my mistakes, but it was a PMR going on and I was giving them ALL information they asked, including model, code, settings, full traces. And they were saying that JHPB is running, but now (when I understand logs more) I see - that it was NEVER started properly.
It was not my homework - I expected them to lead me in my problem solution and I didn't get help I expected to get.
Actually I had several PMRs with IBM and I was VERY VERY pleased to deal with their support. Really. It was my first bad expirience with them. |
|
| Back to top |
|
|
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
|
|
