| Author |
Message
|
| abhyyy |
 Posted: Mon Jun 10, 2013 9:17 pm Post subject: Deciding when and which Node to use ? |
 |
|
Voyager
Joined: 29 Sep 2011
Posts: 83
|
Hi Friends,
I am looking for information on Which node to use or consider while designing a flow and what should be deciding factors. Appreciate if someone can share knowledge on this.
Which node to use and when
-SOAP node VS HTTP node
-Compute Node
-Mapping Node (Choosing over Compute node)
-XSLT node.
-Filter node (vs routing / routetolabel nodes)
-Database node (Choosing over compute node)
-MQget / MQinput (for starting node in flow)
Many Thanks in advance. 
_________________
----------------------
NeVeR StOp LeaRnInG. |
|
| Back to top |
|
 |
| manoj5007 |
| Posted: Mon Jun 10, 2013 10:40 pm Post subject: Re:Infocenter |
|
|
Acolyte
Joined: 15 May 2013
Posts: 64
|
| For all the details you had asked for, please refer to the IBM Infocenter. |
|
| Back to top |
|
|
| abhyyy |
| Posted: Mon Jun 10, 2013 11:59 pm Post subject: |
|
|
Voyager
Joined: 29 Sep 2011
Posts: 83
|
The kind of info I am looking for is :
1.When to choose Mapping over Compute node when u have some transformation requirements ?
2. When to choose SOAP over HTTP or vice versa when u have to go for Sync/async webserives communication between client server?
Infocenter doen't have this info !!
_________________
----------------------
NeVeR StOp LeaRnInG. |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Tue Jun 11, 2013 3:19 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| abhyyy wrote: |
The kind of info I am looking for is :
1.When to choose Mapping over Compute node when u have some transformation requirements ?
2. When to choose SOAP over HTTP or vice versa when u have to go for Sync/async webserives communication between client server?
Infocenter doen't have this info !! |
The InfoCentre does have this info, so do the IBM Training classes. This is basic stuff.
Mapping node is useful when your mapping like structures, such as XML to XML. It's less convenient to map unlike structures in mapping node, as in when the data format needs to change in XML structures.
SOAP is useful over HTTP when you have a WSDL. If you have no WSDL, don't use SOAP.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| mqjeff |
| Posted: Tue Jun 11, 2013 3:19 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
| abhyyy wrote: |
The kind of info I am looking for is :
1.When to choose Mapping over Compute node when u have some transformation requirements ?
2. When to choose SOAP over HTTP or vice versa when u have to go for Sync/async webserives communication between client server?
Infocenter doen't have this info !! |
Because it's not the kind of thing that can be decided generically.
It's site and usecase specific, and you must make a decision based on your knowledge and intelligence. |
|
| Back to top |
|
|
| mqsiuser |
| Posted: Tue Jun 11, 2013 4:37 am Post subject: Re: Deciding when and which Node to use ? |
|
|
Yatiri
Joined: 15 Apr 2008
Posts: 637
Location: Germany
|
There are nodes that do a specific task, I guess you should use that node for the respective specific task  :
SOAP: WebServices (fall-back to an "underlying" protocol (e.g. HTTP, JMS) may be an option though)
HTTP: HTTP-Requests
MQGet: For getting a message when you are in a flow
MQinput: For getting a message from a queue and starting a flow
Then there are 5 transformation nodes, which all do the same thing (transformation)... but also various other things. Some things are only possible e.g. in Java, other only in ESQL... that may influence your decision.
| manoj5007 wrote: |
| For all the details you had asked for, please refer to the IBM Infocenter. |
For transformation IBM says that there are different scenarios (where one node is better suited than the other), but there is at least one other person (in this forum) complaining that they don't give clear enough advice(s) when to use which.
Though mqjeff is right on site-specific (e.g. your customer or fellow collegues love the W3C) and use-case-specific (e.g. simple transformations)... and always think and put in all your intelligence! ... and still ... here I try to capture the knowledge of the experienced developers mind set.
_________________
Just use REFERENCEs |
|
| Back to top |
|
|
| Vitor |
| Posted: Tue Jun 11, 2013 4:40 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| mqjeff wrote: |
| It's site and usecase specific, and you must make a decision based on your knowledge and intelligence. |
There are a thousand points to consider in any design which make the kind of "decision tree" you're looking for impossible.
Consider Mapping v Compute node for transformation. A general rule (for example) could be that Mapping is used for simple transformations and Compute for complex ones.
Immediately you get bogged down in the definition of "simple". Ask 5 people here what a simple transformation is and you'll get 5 answers.
And if you have a "simple" transformation for an XML to which you have no XSD the Mapping node is unusable.
And why a Compute node? If you have a predominently Java shop a JCN is more appropriate.
And what about all the other transformation options WMB offers? Why not them?
Etc...
Etc...
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| mqsiuser |
| Posted: Wed Jun 12, 2013 12:36 am Post subject: Re: Deciding when and which Node to use ? |
|
|
Yatiri
Joined: 15 Apr 2008
Posts: 637
Location: Germany
|
| abhyyy wrote: |
| -Database node (Choosing over compute node) |
I would always use an ESQL-Compute node (and within that likely PASSTHRU).
| abhyyy wrote: |
| -Filter node (vs routing / routetolabel nodes) |
I would also prefer an ESQL-Compute node over a Filter-Node and doing a filtering in ESQL... Probably a filter node can (in some cases) make it clearer (than a compute node) that "just filtering" is happening.
Regarding the routing-nodes... routing based on a db-table and using the "destination list" (within the MQOutput-Node) can help making the routing more configurable/adjustable (in the DB-Table). But if you don't need that a route to lable node is certainly useful.
_________________
Just use REFERENCEs |
|
| Back to top |
|
|
| nathanw |
| Posted: Wed Jun 12, 2013 1:10 am Post subject: |
|
|
Knight
Joined: 14 Jul 2004
Posts: 550
|
Simple answer
You look at the info you have and use knowledge, gathered or acquired to decide
Complex Answer
Gain the knowledge buy actually building flows using different nodes and find out which is best
This does many things it helps you gather the knowledge you require and therefore become a better person for it.
It also allows you to then answer your own questions and learn while you do.
Nothing worthwhile is ever easily acquired
_________________
Who is General Failure and why is he reading my hard drive?
Artificial Intelligence stands no chance against Natural Stupidity.
Only the User Trace Speaks The Truth  |
|
| Back to top |
|
|
|
|
