| Author |
Message
|
| me@tcs |
 Posted: Thu Mar 25, 2010 9:36 am Post subject: Invoking WTX 8.2 maps from WMB |
 |
|
Novice
Joined: 22 Mar 2008
Posts: 18
|
Hi,
My requirement is to invoke maps built in WTX 8.2 from WMB message flows using the WTX map node. I am using type trees generated from XML schemas, with validation option set as Xerces parser, as the input and output cards in the TX map. I have all the required files in the project folder as required by the Xerces validation. Would there be any significant impact of using Xerces parsing option as opposed to Classic when invoking maps from Message Broker? Any help greatly appreciated.
I am using WMB 6.1.
Thanks in Advance. |
|
| Back to top |
|
 |
| Esa |
| Posted: Thu Mar 25, 2010 11:46 am Post subject: |
|
|
Grand Master
Joined: 22 May 2008
Posts: 1387
Location: Finland
|
Hello you@tcs,
I don't think there will be any significant difference in performance impact. I guess xerces might perform slightly better. Anyhow, if I were in your place I would select Xerces, it is easier to map and - if I have understood correctly- the recommended way to map XML with WTX now |
|
| Back to top |
|
|
| me@tcs |
| Posted: Thu Mar 25, 2010 12:31 pm Post subject: |
|
|
Novice
Joined: 22 Mar 2008
Posts: 18
|
| Yeah i understand that using Xerces is the trend right now but as far as I understood and as per my seniors at work, WMB 6.1 in conjunction with WTX map node has a drwaback when it comes to usage of Xerces parser while generating type tree. Looks like output card settings of maps built with type tree having Xerces parser cannot be overridden at runtime. Was wondering if have any info about that. Thanks! |
|
| Back to top |
|
|
| Mercator |
| Posted: Wed Mar 31, 2010 8:45 am Post subject: |
|
|
Apprentice
Joined: 21 Jul 2009
Posts: 34
|
Why build type trees off of XML schemas? Use the schemas natively for input / output cards.
If you have XML schemas and broker is passing the XML data in to WTX, don't use either *.mtt or *.xsd for the input card. The WTX node has the ability to use Message Sets (*.mxsd) natively for an input card. Doing this causes WTX to access the broker elements directly. Nothing is serialized and subsequently reparsed = higher performing. |
|
| Back to top |
|
|
| ghoshly |
| Posted: Thu Apr 01, 2010 2:35 am Post subject: WTX node's ability to use Message Sets (*.mxsd) |
|
|
Partisan
Joined: 10 Jan 2008
Posts: 333
|
Hi Mercator,
I am working with WTX node with XML input, can you please explain how
WTX node has the ability to use Message Sets (*.mxsd) natively for an input card ? what settings need to be done.
Waiting your response. |
|
| Back to top |
|
|
| Mercator |
| Posted: Thu Apr 01, 2010 6:52 am Post subject: Re: WTX node's ability to use Message Sets (*.mxsd) |
|
|
Apprentice
Joined: 21 Jul 2009
Posts: 34
|
| ghoshly wrote: |
Hi Mercator,
I am working with WTX node with XML input, can you please explain how
WTX node has the ability to use Message Sets (*.mxsd) natively for an input card ? what settings need to be done.
Waiting your response. |
Its very simple. Just point your input card to the *.mxsd of your message set. When you are selecting the "Type" on the input card, you must choose the top level, which is "XSD". If you choose anything else, it will not work correctly. Thats all there is to it.
This assumes that broker will be parsing the data using this message set prior to passing the data to WTX. You cannot pass a blob to the WTX node when using the message set in place of a type tree. In your case, the message you are passing should already be in the XMLNSC domain using the same message set.
The only negative aspect to using *.mxsds in place of type trees is that you CANNOT test or run your map standalone. It must always be invoked by broker, even for simple testing. However, its not all the time consuming to create a two node flow (MQInput -> WTX node) to test your map. |
|
| Back to top |
|
|
|
|
