As the tool’s name suggests, it’s a ETL oriented tool. So it allows to perform extraction, trasformation and load. If i’ve understood correctly, these operations are “Data Warehouse” oriented.
is ETL a type of Data Integration?
I’m confused because in my opinion data integration is a totally different thing.
Data integration integrates different data sources and provides a unique global view of these sources (GAV approach).
In ETL doesn’t seem to me that we get a global view of the sources. The “final result” of an ETL process can be very different from the sources. I don’t even know if we can talk about global view, GAV or LAV…
In general, yes. ETL can be seen as a part of data Integration.
Data integration is used in a wider term to connect two or more data sources/targets. ETL is then a technique how to easily connect to these sources, transform the data (if needed) and load to the target. The advantage is that ETL process works as a pipeline so that big amounts of data can be processed.
The GAV/LAV/unified views are just the way how you design your data integration/etl processes. You would use CloverETL for that. You can transform/unify the data or just mediate them – literary anything as it is up to you to architect what you actually want to achieve. The tool enables you to solve it quickly and efficiently.
For further info, I suggest going through some wiki articles explaining these terms, including nice figures. If you decide to try it out practically, feel free to check the CloverETL Designer that you can download from http://www.cloveretl.com/download
So… GAV or LAV (also GLAV), depend on how i build the graph? Isn’t it a thing that depends on the tool itself? Anyway… from what i know, actually the LAV approach is not possible.
If i have understood correctly, i would say that with this tool is possible to obtain a global schema as view of the sources (GAV approach), in which the data at the sources are always materialized (data are always moved from the sources to the final database).
I’ve also read the “data exchange” process. It seems to me that this tool it’s even more similar to the “data exchange” process… or as we already said, it depends on how we build the process.
GAV or LAV (also GLAV), depend on how i build the graph? Isn’t it a thing that depends on the tool itself? Anyway… from what i know, actually the LAV approach is not possible.
It is indeed a matter how you build the graph. CloverETL can prepare the data for you and it is up to you where the data ends up.