Loose coupling transport in PI 7.0
Hi,
We are trying to implemet OTO.I am tring to do a loose coupling transport in PI 7.0 from DEV to QUA. I have created one workbench request in se09 and have exported my IR object. I can see my exported IR object in IR>Tools>Find transports. But Open transport organiser is disabled. Can anyone tell me why is it disabled and how to resolved it?
I feel configuration is not proper. Can any one tell me, where the configuration would be wrong.
Please also send me step by step procedures (links) for implementing OTO
Thanks,
Arivarasu S.
Hi,
PI scenario is just used for XI
Regards,
michal
Similar Messages
-
Hi All,
Sorry to ask this question.but i need exact definition and adavantage of loosely coupled system to undersatnd client the usage of XI in this concept.
One more question same like above.
They are asking interdependency of exchanges , when exists and if not implemented through XI may cause error handling very difficult, how can u achive this in XI.
Though i have some view on the above questions I'm expecting live definitions and usage of above advantages.
Regards,
venu.Hi Venu,
Its better if you think of the concept of loosely coupled systems in a "Before" & "After" kind of approach. To make it simple,
1. System A & System B are a part of your integration landscape.
2. Before introduction of XI (for that matter any similar functionality product), these system integrate with each other making peer to peer connections.Say System A copies a file abc.txt into a folder of System B and vice versa for response messages if any.
3. After introduction of XI, you now leverage onto the integration capabilities of XI. So system A which had to handle copying of file abc.txt to some folder of System B now just has to ensure that it makes the file available to XI which in turn will connect to System B and send appropriate data in desired format.
In this way the two systems which were tightly integrated with each other now are loosely coupled with each other.XI can also take care of any specific requirements of polling periodically , format conversion and many other.
Rgds,
Amol -
Mapping Template : Loosely Coupled or hard Coupled
Hi All
I m using mapping template for graphical mappng. I have the following doubts:-
1. Whether the mapping template is loosely coupled or tight coupled , i mean whether it picks the template at run time or the fields are filled one time only.
2. If we use the mapping template and afterwards do some changes in the mapping template then whether the changes gets reflected in the mapping program or not?
3. Can we change some fields of the mapping templates used in the mapping program after impoting that template? ( Here changes means changes in the mapping program )
Regards
Amithi,
>> Whether the mapping template is loosely coupled or tight coupled , i mean whether it picks the template at run time or the fields are filled one time only.
i dnt know whether this is a standard term or not but in these terms i think it is " tight" coupled.
>> If we use the mapping template and afterwards do some changes in the mapping template then whether the changes gets reflected in the mapping program or not?
No changes dont get reflected automatically if u change the template after assigning it in a mapping program.
>> Can we change some fields of the mapping templates used in the mapping program after impoting that template
It becomes a complete separate mapping program altogether even if u use a template. template is only to save ur time of designing mapping it does not contain references at run time. u can change this mapping program at anytime u want.
[reward if helpful]
regards,
latika. -
How to create loosely coupled transaction branches with DBMS_XA ?
Hello,
I use DBMS_XA to start several sessions in the same database instance participating a global transactions. The local transaction branches are tightly coupled.
Do you know a way (Oracle 11.2) for the branches to be loosely coupled by using dbms_xa in one-only instance ?
Best Regards,
GregoryFor non XA datasource create a JDBC Connection with data source class
oracle.jdbc.pool.OracleDataSource
For xa datasource create a JDBC connection with data source class
oracle.jdbc.xa.client.OracleXADataSource -
Attemp to build loosely coupled WS using xsd:any & xsd:anyType?
Dear All,
i want to write a loosely coupled web service that allow
the server to evolve independently to the client. (main
purpose is to extend the service by adding more properties
in the request / response message)
i want to achieve the loosely coupled goal by adding
xsd:any or xsd:anyType to the request / response messages
(i.e. wsdl ComplexType). (just follow the normal XML
guideline).
i have try to write simple program with xsd:any (which
mapped to SOAPElement in JAX-RPC). And the result seems
good and can achieve my goal. ( i can consume the service
in java and .net without problem).
but later, i read some article, in the web, (like this: http://www.ibm.com/developerworks/xml/library/ws-tip-xsdcaution.html)
that clams it is better to use xsd:anyType instead of
xsd:any (as anyType is mapped to SOAPElement or java
object if server/client can deserialize it).
as far as i know, the support of xsd:anyType is not standardized / madatory.
could anyone tell me your comment about using xsd:anyType or xsd:any?
how about the support of xsd:anyType across different platform / tools?
xsd:anyAttribute seems not supported by jax-rpc, could you please comment on using this element in the wsdl? good or bad?
oracle will map this element to an oracle proprietary java object.
am i choosing the right way to achieve loosely-coupled WS? any other way?
some article claims to send request as xml doc. but it
seems that is not very convenient to server and client as
both sides need to write code for serialization /
deserialization.
why not use the existing automatic XML <--> java object provided by the jax-rpc standard?
Thank you very much
Message was edited by:
lspHello,
Is your question related to very specific element of your message or for the complete message.
Using any/anyType even if powerful could be sometimes complex for Web Service Client developer to handle since the Java object (SOAPElement or Object) is not typed so doing late binding of schema force people to do the mapping themself any.
So before choosing this in a wide scope on your project think also to its impact for end user/developers, and I think you are on the good path, as you said anyType will recognize the object.
I am inviting you to also look at the Oracle Web Services Interoperability documentation.
and this article from SUN: Interoperability With Patterns and Strategies for Document-Based Web Services
Regards
Tugdual Grall -
Java data types are tightly coupled or loosely coupled ?
java data types are tightly coupled or loosely coupled ?
Is this another interview question? If so, the answer you should have given is, "You, sir, have been smoking too much crack."
-
Are WTC transactions loosely coupled?
Hello,
I am looking for a confirmation that XA transactions between Tuxedo and WLS (using WTC) are only loosely coupled.
I cannot find much details on how WTC handles transactions from and to Tuxedo.
On the Tuxedo side, it is stated clearly that transactions accross domains are loosely coupled. Since Tuxedo and WLS instances are in different domains, I am assuming that transactions between Tuxedo and WLS have the same constraints as Tuxedo inter-domain transactions. Unless WTC were to optimize such transactions.
The specific challenge we face is as follows (using Tuxedo 10.0, WLS 10.3, Oracle 11g):
1) A Tuxedo Corba App starts a transaction and does some work.
2) The Tuxedo Corba App calls an EJB service in WLS.
3) The EJB service does some work and updates a record in the Oracle DB.
4) The EJB service returns control to the Tuxedo Corba App.
5) The Tuxedo Corba App selects the same record that the EJB updated.
6) The Tuxedo Corba App does some more work, and the commits the transaction.
The problem is that in step (5) we don't see the update that the EJB made. If we replace the select in step (5) with an update, the update times out because the record is locked. However, the global transaction commits or rolls back correctly on both sides. This indicates that the transaction branches in Tuxedo and WLS are loosely coupled.
RogerAcross Tuxedo /Domain the transaction is loosely coupled. WTC is subjected to the same limitation. So you are correct, those XA transactions between Tuxedo and WLS are loosely coupled.
-
Tightly and Loosely coupled scenarios
Hi Experts,
Could you please provide a example for Tightly coupled and Loosely coupled scenarios from SAP Platform?
How to do the below scenarios from SAP platform?
1. Tightly coupled (Non-SOA):
Business Services : I have to create RPC encoded style webservices.
Business Process : I have to call other platform RPC encoded style webservices (More than 2 webservices).
2. Loosley coupled (SOA):
Business Service : I have to create Document literal style webserivces
Business Process : I have to call other platform Document literal style webservices ( More than 2 webservices)
Please provide the input to do the above scenarios from SAP platform and please inform me which tool/component I have to use do the scenarios.
Thanks & Regards
SaraHi Sara,
Can you please elaborate your question in detail .
1) Web Services are always examples of loosely couples architecture, either its document style or RPC style and whether you are creating WS or consuming WS.
Tightly coupled architectures were those whose business logic layer and presentation layer were tightly coupled and reusability or scalability was a problem
If you have user interfaces involved in your business process, then you should go for CAF( Composite Application Framework) to define a business process using VC, WebDynPro or Adobe Forms as UI.
If it doesn't involve any user interaction, then you can go for ccBPM of Netweaver-PI to automate the process.
Regards,
Piyush -
Tightly coupled and Loosely coupled scenarios
Hi Experts,
Could you please provide a example for Tightly coupled and Loosely coupled scenarios from SAP Platform?
How to do the below scenarios from SAP platform?
1. Tightly coupled (Non-SOA):
Business Services : I have to create RPC encoded style webservices.
Business Process : I have to call other platform RPC encoded style webservices (More than 2 webservices).
2. Loosley coupled (SOA):
Business Service : I have to create Document literal style webserivces
Business Process : I have to call other platform Document literal style webservices ( More than 2 webservices)
Please provide the input to do the above scenarios from SAP platform and please inform me which tool/component I have to use do the scenarios.
Thanks & Regards
SaraHi Sara,
Can you please elaborate your question in detail .
1) Web Services are always examples of loosely couples architecture, either its document style or RPC style and whether you are creating WS or consuming WS.
Tightly coupled architectures were those whose business logic layer and presentation layer were tightly coupled and reusability or scalability was a problem
If you have user interfaces involved in your business process, then you should go for CAF( Composite Application Framework) to define a business process using VC, WebDynPro or Adobe Forms as UI.
If it doesn't involve any user interaction, then you can go for ccBPM of Netweaver-PI to automate the process.
Regards,
Piyush -
Using a Singleton Mediator to provide loose coupling
I need to discuss the pros and cons of using a singleton mediator to provide loose coupling between component implementation classes.
Suppose you are building a SE or a BC where an endpoint is defined upon a SU deployment. In this scenario, you will end up finding out that the Component, ComponentLifeCycle and ServiceUnitManager implementations should know each others instances. I can figure out these solutions:
- Use tight coupling
- The component implements the three interfaces (or two of them and associates the third)
- Use a singleton mediator to hold the instances
- Use a third party solution, like Spring's ApplicationContext and ApplicationContextAware contracts
- Some other solution I didn't think of (the scenario has to be more specific)
Assuming the scenario as described above, I prefer the third solution.
I appreciate your opinions.
Hossam KarimWhen you evaluate the JBI implementations available, you will notice that each one has built a less abstraction layer over the JBI component contracts in order to simplify its subsequent tasks. As a component builder, you would either reuse one of these layers and get tight with a specific implementation, or start writing your own.
I recently proposed a simple framework to my team; you can find a class diagram here (http://www.gaiati.com/products/emee/framework.png), it has the following elements:
� AbstractBootstrapAdapter: Simple adapter for the Bootstrap interface
� ComponentApplicationContext: Provides a Spring based Application context for this component, the context beans should be loaded upon component initialization. An instance of this class will be available to all classes or interfaces that extend the ApplicationContextAware Spring interface. All classes and interfaces on the previous diagram that associate a ComponentApplicationContext instance (applicationContext) implement the ApplicationContextAware interface. One exception is the AbstractComponentAdapter class, because it actually creates the instance.
� AbstractComponentAdapter: Adapter for the component and lifecycle contracts. Upon initialization, this class creates a ComponentApplicationContext instance, and uses the beans to configure itself, the class recognizes other framework contracts and able to communicate with them.
� AbstractMessageListener: Consumes this component's DeliveryChannel message exchanges, and fires a handling trigger.
� DeployedEndpoint: Represents an endpoint that was deployed using a SU
� DeployedEndpointConfiguration: Holds the configuration needed by a DeployedEndpoint instance
� DeployedEndpointTarget: Represents the target JBI specific address that a DeployedEndpoint instance should route the messages to.
� GenericMessageListener: Handles the trigger fired by the AbstractMessageListener, delegating the work to an instance of the DeployedEndpoint interface
� GenericServiceUnitManager: Handles GenericServiceUnit instances
� GenericServiceUnit: Acts as a proxy and encapsulates a DeployedEndpoint instance
� AbstractDeployedEndpointFactory: An abstract factory for creating DeployedEndpoint instances using DeployedEndpointConfiguration instances
Yes, I ended up using Spring, and you are right, it might not be appropriate to attempt to separate the interfaces, may be I misunderstood or was mislead by Open-ESB implementation of the SOAP and file binding components.
It is true sometimes you can have "personal taste", but most important is to achieve best practice.
Finally, I hope the specification in its next versions, would provide us with a best practice framework or proposal, instead of so much choices.
Hossam Karim -
Hey guys,
I have 3 files: main.mxml, registration.mxml, and
update.mxml. The registration is a custom canvas component with a
datagrid, and the update is a popup activated by clicking the
datagrid. I understand how to dispatch custom events, but I'm
having trouble dispatching from popup back to main.mxml where my
remote code is located. I can successfully dispatch objects with a
custom event from reg component to main. What am I doing wrong?
Should I just add a remote command in the registration component
and run a function in the registration? Doesn't seem like loose
coupling to me?If your registration component is instantiated in main and
you dispatch an event that bubbles from registration to main, then
bubbling works. But forget about bubbling when dispatching an event
from a popup to any other component. You can do the following:
If you use a model (with Cairngorm, PureMVC ... or simply
Singleton), you can have an object in your model that extends
EventDispatcher and serves as a message target.
package com.mycompany.core.model
import flash.events.EventDispatcher;
// This class only serves as the target for messages that
concern objects that are not
// in the display list or that are in the display list but
cannot be reached by event bubbling
public class MessageTarget extends EventDispatcher
public function MessageTarget()
In your model:
// Message target
public var messageTarget:MessageTarget = new
MessageTarget();
In any component interested in some messages (e.g column
visibility) dispatched on the message target :
// Add event listener for column visibility change
this.model.messageTarget.addEventListener(ColumnVisibilityChangedEvent.COLUMN_VISIBILITY_ CHANGED,
this.onColumnVisibilityChanged);
Then, you can dispatch an event from anywhere that can be
listened to anywhere:
// Send Column Visibility Changed event
this.model.messageTarget.dispatchEvent(
new
ColumnVisibilityChangedEvent(ColumnVisibilityChangedEvent.COLUMN_VISIBILITY_CHANGED,
false, false)); -
Loose coupling among objects within oracle schema
I am building an information service that manages Suppliers. The suppliers are used by our billing system, tender system and sales system. Though 60% of the attributes of supplier are unique to each system, there are still 40% attributes of Supplier that are shared across the systems.
My objective is to build a flexible system, so that change to one individual system's data, should not impact other systems. For example, if i need to make certain tables offline for upgrading them, it should not impact rest of the systems that need supplier information. What is the best way of achieving this? Should all the different context specific attributes live in one schema, but deployed on different table spaces? Also, the read and update may happen more for one set of attributes than the other. How should i logically represent them via one model, but deploy them in such a fashion that they can evolve independently?
Thank you.Reports - All reports - Data dictionary reports - Database administration - All tables - All tables, and specify B as schema name.
Have fun,
K. -
Loose coupling of metadata...how?
Hi,
I have database AAA,BBB,CCC
AAA-Source
BBB- Development Database with OWB repository(OWB is installed here)
CCC-Target database(doesnot have any OWB installation)
Now I have created a set of mappings in BBB to populate tables.i.e the logical design is in BBB.
Now I want to run the same set of mappings existing in BBB to populate the target tables in CCC.
Can I do this without actually installing OWB in CCC.If so,can you please let me know how to achieve this.
If not,how should I proceed.
Its an urgent help I request you for.Please respond.
Thanks Much
SridaHii,
Appreciate your quick reply.
Sorry for being not clear in the earlier post.
AAA,BBB,CCC are different databases,they are not schemas
AAA is source which I am remotely connecting to....BBB is the database on my local Machine with OWB installed and ETL defined...again CCC is a remote DB.
i.e I am trying to pull the data from source DB and do necessary transformations locally and Push it to target DB.
Can you guide me as I am newbie.
Thanks Much!
Bregards
Srida -
Can OpenMQ use SMTP as transport in a distributed setup?
Hello all,
I am researching the usability of OpenMQ to provide messaging services to a number of J2EE
servers and clients distributed in a multi-segmented network (like central office and branch offices).
Deployment in each network segment is complete and looks like any other segment's deployment -
they all have GlassFish app servers, database and LDAP replicas, etc. We plan to connect some
GlassFish-deployed applications and user-interface clients with MQ services, with a broker and
each queue's instance deployed in each network segment.
So my first short question is: am I correct to assume that in order for OpenMQ instances to
pass messages between clients in different network segments, all of these instances should
be pre-configured as a single "conventional cluster"? Is this the only way? That is, there is no
such loose coupling as in email services "I've got a (spam) message for a random recipient
name, try to deliver it please"?
Second, can this be implemented as a multi-tier cluster? That is, "home brokers" in different
network segments are inter-connected as a conventional cluster (brokers with dedicated
datastores), but in fact these "home brokers" within a network segment are implemented
as an active-active cluster with a shared datastore (local to the segment). Can this be done?
Third, hard and titular question is: some of the network segments have no online IP
connection per system design, they are physically separated. There is a means of
SMTP transport between nodes in these segments, using offline techniques, though.
Can OpenMQ nodes exchange messages (client traffic and/or administrative traffic
like cluster config propagation), using SMTP/POP3/IMAP instead of online connections
such as JMS/TCP or HTTP?
I have not found any indication of this approach in the documentation, except that GlassFish
appservers seem to be able to implement JAX-WS over SMTP as experimentally supported.
Thanks,
//JimHamper wrote:
I was wondering then, as I don't have a Time Capsule, whether it would be possible to dedicate the second HDD for exclusive use of Time Machine, backing everything up?
Anyway, would splitting the two work and would it be easily achievable with existing tools in OS X?
Yes, you can use the 2nd HDD with Time Machine as your TM back up drive.
As long as you are able to format your new SSD and the HDD, you can use them independently. I would advise using Terminal and the diskutil commands to do this as Mountain Lion 10.8.3's GUI Disk Utility will see the SSD and HDD as needing to be "fixed" (drives marked in RED) and if you choose to FIX the drives your two drives will automatically be setup as a Fusion drive and all data on the drives will be lost. -
Discussion: File CC Transport. Setting location of Folder, Server, Archive
Dear Experts,
I want to open a discussion.
This is not something technical but rather to be more managerial.
- When we transport the objects in Integration Builder to another server (like from DEV to QA), who do you think should be responsible to activate the transported objects? Will it be the basis guy or the development guy?
If the Basis how can they find which object to be activated easily? Is it from the change list? Because the Basis guy usually does not know about the objects, he is just given the TR and then transports it.
- This is a similar question but this is related to the File Communication Channel. Because when we transport the file communication channel we need to re-configure the filename, location, archive location, FTP server, username and password (for object has not been transported before).
Who do you think should re-configure that?
Any opinion is appreciated.
Thank you,
Suwandi C.Hi,
I use the file transport mechanism.
I've been trying to configure the CTS+, but fails as I've read from the following document:
http://www.sdn.sap.com/irj/scn/go/portal/prtroot/docs/library/uuid/20ccd0ec-2839-2b10-0097-
828fcd8ed809?overridelayout=true
In Chapter 2 it is stated that
"Please keep in mind that the CTS+ Close Coupling integration is only available for SAP NetWeaver
Process Integration 7.1 starting of SP06.
If you have older Patch Levels, only the loose coupling via File Export and upload of Non-ABAP
objects to a CTS+ communication system exists. "
Since I'm using the PI 7.1 SP04, I guess I have no other choice but to use the File transport mechanism.
So then, I have no choice instead of having the basis guy to activate and configure CC when transporting the PI Objects in Integration Builder.
EDIT:
I've just doing some reseach and it seems that only the close couping for SAP PI available starting at version 7.1 SP06, For older version the loose coupling CTS by uploading the non-ABAP object to the TR is still possible. Is this correct?
https://help.sap.com/saphelp_ctsplug20sm71/helpdata/en/09/734a27116b4f6185e072df4e8332fa/content.htm
https://help.sap.com/saphelp_ctsplug20sm71/helpdata/en/f5/6a5c15357b4db19fb07f5ee1a97472/content.htm
If this is correct, I need more insights of how this work. Will it work like this:
PI Developer will chose the PI Objects to be transported and export the objects to a *.tpz file (using the file system mechanism). PI Developer will then attach the file system to a TR from the ABAP TCode SE09. TR will be given to the Basis guy to be transported. Once Transported the PI Objects in Integration Directory will be based by the Developer user ID because it is using the CTS mechanism transport.
is that correct?
Thank you,
Suwandi C.
Maybe you are looking for
-
How to let mmp_smtp support virtual domain authentication by user/pass
my mmp had configured 2 virtual domain ,one is mmp1.soft.com,other is mmp2.soft.com i can send mail by mmp use the user/pass format : mmp1user/pass , which is sotred in o=mmp1.soft.com,o=isp in ldap,and the format [email protected]/pass which is stor
-
How to have more than one free good for a material ?
Hi everyone, we have this requirement from the user where he wants give more than one free good for a material. As far as i know we can configure only one free good for a material in standard SAP. if have any idea about how to achieve this please gui
-
Need Help on Data Back up and Retrieve
Hi Mac Helpers I have a Macbook Pro Late 2008 Model and have both my Iphone and iPad synced to it. I intend to save all my data from the macbook to a Time Machine and then sell my macbook as I intend to buy a new one this July. My questions are 1) wo
-
PDF file size grows each time form is saved (with no data changes). Why?
I have an SAP Adobe form (PDF) and we are seeing a difference in the Save behavior between when it is viewed in Adobe Reader vs. the adobe plug-in(?) for Internet Explorer. When the form is viewed in Internet Explorer, every time the user saves the f
-
Hi all, i am trying to create a new page layout with required structure. Please help me on the steps to be followed. Thanks, Gowt