One Star

[User scenario] ESB or Data Integration?

Hi everyone,
I am new to Talend but have just worked through the first 2 ODT modules of Talend.
Now I want to put it to use and create a simple application which works like this:
1. It receives 3 simple CSV files with 3-5 columns. Really simple stuff.
2. It provides a SOAP Web Service which can be called from other applications
3. Such application would only provide 2 request parameters: a selector and a customer number. E.g.= Selector A and customer number 4711
4. Based on the selector, the service chooses one of the 3 files as data source
5. It then searches within that file (or a database table representing that file.) for records matching that customer number
6. If no matches are found, a fault message is returned
7. If at least 1 match was found, the whole result set is returned as sequence of a complex type. Essentially just a list of records that have like 3 elements each.
So you see: it almost doesn't get easier than this.
But since I am a newbie, here are my questions for you:
1. Would you choose Talend ESB or Talend Data Integration for this?
2. Why?
3. How would you go about this?
4. Can such service be configured in a way that it uses different URL's depending on the environment (Test, Production?) Or rather: can that be achieved by using contexts?
5. Can such service implement certificate based authentication? The reason is that the actual partner system will be a real Enterprise Service Bus which requires that we exchange certificates to establish mutual trust.
Thank you very much
Seventeen Stars

Re: [User scenario] ESB or Data Integration?

This use case can be done with the Data Integration capabilities. From my point of view the DI stuff is much easier to maintain and to understand as the ESB stuff.
I would definitely start with DI!
One Star

Re: [User scenario] ESB or Data Integration?

Thanks for your helpful reply. One further question, though: how do I implement #4? (Choosing the file for the search based on the selector?) Essentially, the files have a different format and might even result in a different response-type. So in essence, I must be able to implement 3 different processes (one for each file) that generates the response - and then choose which path I take.
No idea how to do that :-( Do you?