Is near-line storage the preferred solution for data aging in BW? Under what circumstances would near-line storage be preferential as a storage medium rather than the more conventional method?
One definite goal of data warehousing is to handle large volumes of historical data. Unlike in ERP systems, the end of life of data or information is not clearly defined in a data warehouse. The older data becomes, the more valuable it can be, such as for data mining purposes, for example. In a conventional archive, infrequently used data and information is too far away for ad hoc queries or load processes (in terms of remodeling, and so on), because archives do not normally allow direct access to their data. There is a certain gap between the database and the archive file that needs to be filled. This is what NLS is about: large amounts of data that resides on ‘particular devices’ with direct SQL-based access. The particular device here can either be a classic archive file for which index structures have been subsequently generated (PBS Solution), or it can be a blade server with column-based compression algorithms (SAND Searchable Archive).
