| Author |
Message
|
| fjb_saper |
 Posted: Mon Aug 29, 2011 6:27 am Post subject: |
 |
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
| Vitor wrote: |
Yes, but only if the JMS application concerned agrees. I believe we've established further up the thread that this application does expect the headers.
But this whole thread is built on assumptions, suppositions and mismatched technologies.... |
This has nothing to do with the JMS application agreeing or not.
The only reason to force an RFH header or V7 properties is if the JMS application expects specific message properties and needs to process according to those...
If there is no need for the message properties usually contained in the JMS RFH usr folder, there is no "NEED" for such a header (RFH) whether the JMS application expects an RFH header or not...
That was what I was trying to convey with my question. 
_________________
MQ & Broker admin |
|
| Back to top |
|
 |
| Vitor |
| Posted: Mon Aug 29, 2011 6:32 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| fjb_saper wrote: |
| If there is no need for the message properties usually contained in the JMS RFH usr folder, there is no "NEED" for such a header (RFH) whether the JMS application expects an RFH header or not... |
Even if there are no usr properties, would not a JMS-compliant application expect the other RFH2 components to be present?
Given your point that if it's not using usr properties, it won't give a tinker's cuss if there happen to be some.
Given also my expereince is of JMS apps that get cross and throw rafts of Java-ish trace if there isn't an RFH2 on the message they read even though there's nothing in there the application is referencing explicitly in written code. And that I didn't know enough to fight the Java people when they said "oh, the header's missing, add the header". And that putting an RFH2 on makes it work.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Mon Aug 29, 2011 2:23 pm Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
| Vitor wrote: |
| fjb_saper wrote: |
| If there is no need for the message properties usually contained in the JMS RFH usr folder, there is no "NEED" for such a header (RFH) whether the JMS application expects an RFH header or not... |
Even if there are no usr properties, would not a JMS-compliant application expect the other RFH2 components to be present?
Given your point that if it's not using usr properties, it won't give a tinker's cuss if there happen to be some.
Given also my experience is of JMS apps that get cross and throw rafts of Java-ish trace if there isn't an RFH2 on the message they read even though there's nothing in there the application is referencing explicitly in written code. And that I didn't know enough to fight the Java people when they said "oh, the header's missing, add the header". And that putting an RFH2 on makes it work. |
OK let's narrow it down.
We are not talking about an ObjectMessage or StreamMessage(JMS to JMS).
We are restricting for all purpose this discussion to TextMessage and BytesMessage types, typically used in JMS to non JMS communication, and vice versa.
Furthermore let's put out the assumption that the message has no customer specified attributes or properties at creation, and that none are being added as enrichment or being needed at consumption time.
For those 2 types JMS doesn't care about the presence of an RFH(2) or not.
The only reason for an RFH would be a non standard JMSCorrelationId (more than 24 bytes) or some property in the usr folder that is needed by the application.
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| francoG |
| Posted: Tue Sep 06, 2011 2:05 am Post subject: |
|
|
Novice
Joined: 18 Aug 2011
Posts: 23
|
Hello Again Vitor, hello fjb_saper
In the last days I read again and again all the documentation written in 4 hand by IBM and the end-committer of the Web Service (let us name it with a conventional name "Zeta").
I also asked some collegue of mine to read this documentation and we are all astonished and puzzled 
anyway....
About the RFH2 Header, in the documentation of the "Zeta"Web service I found the assert that you MUST write the specific RFH2 Header to communicate with the system. The sample code has been written for this "Zeta"system: I don't think sombody wrote this only for fun.... 
the only strange thing is: why I should write a Java class to simulate JMS RFH2 Header, only to use native MQ ? Why I don't use JMS directly?
I will have a meeting with the "Zeta" Web service in the Next day so I will ask them what exactely have I to do. I hope I will clarify all of these strange things.
About developing in Delphi: I see the Activex MQAX200 and I imported it to Delphi, I finally found the MQQueue, MQMessage Class I was looking for. But Will I use it? 
You(Vitor) told me that is not good choice, despite MQAx200 is still distributed in all SDK for MQ it seems it has given up by IBM.
So I can only trust you!
But at this point I have no other choice: too much things are missing if I use Delphi. 
Sad to say, but I think we will build a sort of JMS Gateway like this:
http://www.ibm.com/developerworks/websphere/library/techarticles/0402_du/0402_du.html
that will forward every request from our application to the "Zeta" web service and give back the result.
I hope this will work in short time..
that's all for now
Thank you very much for your time 
Franco |
|
| Back to top |
|
|
| Vitor |
| Posted: Tue Sep 06, 2011 4:51 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| francoG wrote: |
| why I should write a Java class to simulate JMS RFH2 Header, only to use native MQ ? Why I don't use JMS directly? |
Beats me - I've been advising the use of JMS since way back in this post! 
I will have a meeting with the "Zeta" Web service in the Next day so I will ask them what exactely have I to do. I hope I will clarify all of these strange things.
I think I mentioned using a gateway as well.
I think you can be confident in this working in the short time period you mention.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| Vitor |
| Posted: Tue Sep 06, 2011 4:59 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| francoG wrote: |
You(Vitor) told me that is not good choice, despite MQAx200 is still distributed in all SDK for MQ it seems it has given up by IBM.
So I can only trust you! |
No you don't have to trust me - you can read it here in the WMQv7 documentation:
| Quote: |
| Support for ActiveX has been stabilized at the WebSphere® MQ Version 6.0 level |
So IBM has not given up on it, and it's still supported. But it's already unable to use all the WMQ facilities and this will get worse as time passes. So it's not a good choice for a new development.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| mqjeff |
| Posted: Tue Sep 06, 2011 5:20 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
| francoG wrote: |
| But at this point I have no other choice: too much things are missing if I use Delphi. |
I repeat my earlier suggestion of looking at using either Windows Communications Framework or the Axis libraries from *within* Delphi to perform SOAP/MQ.
I can't say explicitly that either are possible or how to do them, but it is entirely worth your time to spend a couple of hours investigating it. |
|
| Back to top |
|
|
| francoG |
| Posted: Wed Sep 07, 2011 12:15 am Post subject: |
|
|
Novice
Joined: 18 Aug 2011
Posts: 23
|
Hi Vitor,
yesterday I asked to a colleague of mine if he can build a gateway following the instruction we already talk about yesterday....i.e. this: http://www.ibm.com/developerworks/websphere/library/techarticles/0402_du/0402_du.html
Unfortunately this example is for Websphere 5 and we are not able to build it for 6.0 or 6.1: any try we made will gave us errors...
It requires to build a websphere 5.0 TestEnvironment : this is present in also WS 5.1 but does not exists in 6.x.
Any other try with different environment gave us a corrupt server as result. 
Do you know if there is somewhere an example for WS 6.0 or 6.1?
Franco |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed Sep 07, 2011 4:49 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| francoG wrote: |
| Do you know if there is somewhere an example for WS 6.0 or 6.1? |
I've no clue; others might know.
Do you have an overriding need to use Websphere? Something this simple wouldn't appear to need it.
I also echo the comments of my most worthy associate regarding the use of WCF; I equally am uncertain if it would work but it does appear to be an option on the face of it.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| francoG |
| Posted: Wed Sep 07, 2011 6:04 am Post subject: |
|
|
Novice
Joined: 18 Aug 2011
Posts: 23
|
| Quote: |
| Do you have an overriding need to use Websphere? |
unfortunately the environment is not up to me.
does any other server( i.e.apache) be able to use IBM-MQ?
because the MUST is the IBM: the server on "zeta" side is WS-IBM the MQ is IBM-MQ....
 |
|
| Back to top |
|
|
| Vitor |
| Posted: Wed Sep 07, 2011 6:11 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| francoG wrote: |
| does any other server( i.e.apache) be able to use IBM-MQ? |
Yes. Websphere, Apache and any other JMS-aware server can use IBM's offering as a JMS provider; I've covered this point above.
I've insufficient experience to talk in detail about the problems you're experiencing, but will point out that for obvious reasons IBM's Websphere Application Server is designed to work with IBM's Websphere MQ providing the messaging layer.... 
(Indeed I believe I'm correct is saying WAS 6 & later's inbuilt JMS support is provided by a neutered queue manager hidden in the WAS product)
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| francoG |
| Posted: Wed Sep 07, 2011 6:14 am Post subject: |
|
|
Novice
Joined: 18 Aug 2011
Posts: 23
|
| indeed: it would not be a bad idea if I could build a gateway in a less expensive environment that WS... |
|
| Back to top |
|
|
|
|
