| Author |
Message
|
| Cubersome |
 Posted: Tue Mar 13, 2012 7:06 am Post subject: Creating Dynamic output terminals for WTX using WMB |
 |
|
Apprentice
Joined: 02 Mar 2012
Posts: 37
|
Hi,
My requirement is to call a existing map which has a certain number of input cards and output cards. How do I dynamically pass the data for the input cards and dynamically generating the output terminals for that WTX node. I don't have an option to create separate message flows for each map. So I have to do it with one generic message flow and overiding the Input and output card settings dynamically through compute node.Please advice. |
|
| Back to top |
|
 |
| lancelotlinc |
| Posted: Tue Mar 13, 2012 7:42 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
|
| Back to top |
|
|
| mqsiuser |
| Posted: Tue Mar 13, 2012 8:06 am Post subject: |
|
|
Yatiri
Joined: 15 Apr 2008
Posts: 637
Location: Germany
|
Can you rewrite your WTX-Maps in ESQL? Seriously... how much efford is it to fix you WTX issue, and how much efford is it to recode the flow in ESQL... you should think about that (for at least a short moment  ).
output cards, input cards... it is astonishing with what (names) vendors come up sometimes... they must have been "playing cards" too much... the IT naming(s) that I know is/are different.
"So I have to do it with one generic message flow": Probably you should not try to put in everything into one flow, but into serveral: One flow for each input/output message (though I must say that I almost don't do that ever , eventhough it 'might be'/is desireable).
There have been discussions about WTX here in the forum that you might want to read... from what I understood: (The efford of) recoding in ESQL is an option.
| Cubersome wrote: |
| overiding the Input and output card settings dynamically through compute node |
There is PROPAGATE (TO OUTPUT TERMINAL 'abc') and "RETURN FALSE" to propagat several (smaller) messages to output-terminals.
_________________
Just use REFERENCEs
Last edited by mqsiuser on Tue Mar 13, 2012 8:20 am; edited 1 time in total |
|
| Back to top |
|
|
| Cubersome |
| Posted: Tue Mar 13, 2012 8:18 am Post subject: |
|
|
Apprentice
Joined: 02 Mar 2012
Posts: 37
|
The versions are
WMB - 7.0.0.1
WTX - 8.3.0.4 |
|
| Back to top |
|
|
| mqsiuser |
| Posted: Tue Mar 13, 2012 8:22 am Post subject: |
|
|
Yatiri
Joined: 15 Apr 2008
Posts: 637
Location: Germany
|
I think he must be thrilled by my suggestions
Ah, great, we know the exact version now... how valuable.
_________________
Just use REFERENCEs |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Tue Mar 13, 2012 8:33 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
|
| Back to top |
|
|
| Ric-Tic |
| Posted: Tue Mar 13, 2012 11:09 am Post subject: |
|
|
Apprentice
Joined: 18 Nov 2003
Posts: 38
Location: Zurich, Switzerland
|
| lancelotlinc wrote: |
| If your bent on WTX, you should be at 8.4.0.1. |
The OP is on WMB v7, not v8
| Quote: |
| Remember WTX is a legacy product and will be phased out. |
No announcement yet
| Quote: |
| Everything you can do with WTX you can do without WTX using ESQL. |
Not exactly correct, unless you pay for services eg SWIFT message sets |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Tue Mar 13, 2012 11:42 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
|
| Back to top |
|
|
| kimbert |
| Posted: Tue Mar 13, 2012 12:35 pm Post subject: |
|
|
Jedi Council
Joined: 29 Jul 2003
Posts: 5542
Location: Southampton
|
Native WMB flows are inherently simpler and easier to maintain than flows that call out to WTX. Usually faster too. But, to put it mildly, there is some arrant nonsense being spouted on this thread. This forum is intended to be a source of reliable information, so I'll give it my best shot:
| Quote: |
| Can you rewrite your WTX-Maps in ESQL? |
WTX maps can be used for any and all of:
- parsing
- writing
- transformation, enrichment etc
- error handling
- [ a fair few other things]
My point is this: you cannot simply replace any WTX node with some ESQL. Not because it would be difficult ( although it might be ) but because ESQL is not the answer to every question. To replace a flow that calls out to a WTX node might involve the creation of
- one or more message sets ( with MRM physical formats ),
- Compute, JavaCompute mapping nodes
- Aggregation/Collector nodes
- Complex error handling structures
- ...
I could go on, but hopefully you get the point.
| Quote: |
| output cards, input cards... it is astonishing with what (names) vendors come up sometimes |
Well, those are well-known terms to anybody who knows the WTX product. To some people, 'XMLNSC' is a complicated name for an XML parser, but the name doesn't cause many problems for WMB users.
| Quote: |
| Everything you can do with WTX you can do without WTX using ESQL. |
There are not many good reasons for developing new messages flows that use WTX, but there are some. And again, 'ESQL' is an inedequate way to describe the native WMB solution, which might involve all aspects of message flow design.
| Quote: |
| Beginning with WMB 7, there are sample industry flows that illustrate how to use ESQL nodes |
Please describe exactly what you mean by the term 'ESQL nodes'. Alternatively, stop using the term. |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Wed Mar 14, 2012 5:11 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| kimbert wrote: |
| Quote: |
| Everything you can do with WTX you can do without WTX using ESQL. |
There are not many good reasons for developing new messages flows that use WTX, but there are some. And again, 'ESQL' is an inedequate way to describe the native WMB solution, which might involve all aspects of message flow design.
| Quote: |
| Beginning with WMB 7, there are sample industry flows that illustrate how to use ESQL nodes |
Please describe exactly what you mean by the term 'ESQL nodes'. Alternatively, stop using the term. |
The term "Compute" seems too broad. If I said ESQL Compute, would you feel that to be more accurate?
ESQL Compute nodes does not a solution make. True. It takes a comprehensive flow design. Also true.
Thats why I posted links to the industry solutions.
In its current form, WTX is an awful product. It fails to support multi-threading, cannot easily be included in automated builds, will not build on Linux platforms, and generally all around pain in the rear to support in production due to its idosyncs. Some of the workarounds suggested by IBM support include only having one WTX map per EG. If you care to defend it, its up to you. But these technical shortcomings are shameful. This is the 21st century. We have terabytes of RAM, trillions of threads, and still can't put more than one WTX map in an EG.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| kimbert |
| Posted: Wed Mar 14, 2012 5:28 am Post subject: |
|
|
Jedi Council
Joined: 29 Jul 2003
Posts: 5542
Location: Southampton
|
| Just to be clear, I'm not defending anything here - I'm just trying to ensure that this forum remains a source of helpful, accurate information. |
|
| Back to top |
|
|
| mqsiuser |
| Posted: Wed Mar 14, 2012 5:33 am Post subject: |
|
|
Yatiri
Joined: 15 Apr 2008
Posts: 637
Location: Germany
|
choice is good
_________________
Just use REFERENCEs |
|
| Back to top |
|
|
|
|
