|
Enterprise Application Integration and Data Warehouse are separate topics and they can be managed in a way that dependence on them is minimized. If we do not consider effort and focus issue, technically both could be done in parallel, unless there is a significant database change expected due to EAI.
EAI, also includes the integration at data level, so that applications can share the data from the other applications. A Data Warehouse is just like an application which wants to share the data from the source systems. Technically speaking, one can use EAI tools for the data exchange (like deploying Tibco for getting the data from the source system and placing it in the staging area). However, for the ETL operations, one may do without EAI tools, because of the:
- Point to point connections between the source systems and staging area.
- High volume of data flow in a short period of time.
- Extraction is not happening on OLTP mode, and few seconds of minutes of delay is acceptable.
- The extraction is happening mostly as is, and transformation and any kind of complex checks are happening during transformation stage.
In other words, we don't have to compulsorily use EAI tools for Extraction functions. |
