Have a question about how ODWEK works as far as returning documents from our system.
We are using a distributed system with a library server and object server.
My understanding of the flow, is that the ODWEK client sends the Login Requests/Search request to the library server. The Library server then sends the hit list back to the client.
After that, the client sends the retrieve request to the Library server.
Our question is about the request. The data that comes back - Two scenarios- Which one is true - or is there a third.
1) The object server sends the data to the library server, and then the library server returns it to the ODWEK client.
2) The object server sends the data directly to the ODWEK client, bypassing the library server.
Option 2 is how it's supposed to work. The original concept for the object servers were that they were physically diverse. Imagine an organization with offices in Boston, Miami, Denver, and Seattle. The Library server is in Denver, but each office loads their data to their local object server. The metadata gets sent to the library server, and the document data stays local to each office. The design is supposed to be that the client pulls the documents from each object server, and not route them through the library server.
Hope that helps. :)
-JD.