SAP Hybris DataHub solution to data migration and data integration.

Data migration is the essential part of any Enterprise software.
Before datahub it was like to create custom integration project to manage the data migration activities in the project.

When i got task to do datahub implementation i come across many question here is the attempt to write answers to few of them hope it will help you as well :

Q.Why datahub ? Business Analyst view ?
Q.Datahub baisic architecture basic flow.
Q.Some features of dataHub
Q.What is feed ? what is pool?

Q.Why datahub ? Business Analyst view ?
Data Hub is designed to overcome barriers related to the import and manipulation of large amounts of data.
Its claims to be very customizable and solution to any data integration scenario.
This claims is based on the fact that data transformation in datahub is independent of external systems and also on many layers it transfer data asynchronously .

Q.Datahub baisic architecture basic flow.


Raw model comes from the source system we can call this as schemaless defination
In the canonical model are created and transformed on the basis of raw items they are representation of domain model, the canonical models are independent of raw and target items.
Then it comes target model which reconciles the canonical items to target system requirements.

Datahub dataflow is divided into three layers :
1) Load 2)Composition 3) Publication

1) Load :
Data load in datahub is a key value pair , in this we will write rawmodel xml file which will have key value pair to map raw to the canonical entities .
I think we have this layer to completely separate the canonical model from raw and target model otherwise there is no need for this layer in my opinion .

2) Compose :
Here the raw items are grouped , the grouping of raw items are arranged by canonicalItemType or by primaryKey if grouping handlers are not present.
After grouping it undergoes to composition on grouped items ( here it handles the localisation or collection for entities as per canonical model )
we can add the composition handlers if required , now after this we can say grouped raw item has been composed into a canonical item .

3) publication :
here canonical item is published to a target system.
During publication, the publication phase uses the target.xml file to transform the canonical data into target compatible output

Q.Some features of dataHub
1) DataHuub runs on standalone server so it can work independent of the Hybris if necessary in some scenarios.
2) DataHub uses SpEL expressions for custom transformations .
3) provides the test environment which doesn’t require to install Hybris .

Q.What is feed ? what is pool?
Datahub enables data segregation using pools and feed.

There is lot to talk about this i will update more information soon.

Leave a Reply

Your email address will not be published. Required fields are marked *