I am working on good retrieval mechanism for big data systems. I am of view that multidimensional structures could be of great use for aggregations and querying . Multidimensional structures as mentioned in theory/products could replace the need of OLAP CUBE.
Here is some data size info residing into data-warehouse and it is approx. around 1 trillion per month and some time even more and having rolling window of 1 month or greater, it depends per client. I used to write reports,sub-reports and calculate aggregates using sub-reports , and also used analytical and analytical aggregate function to accomplish the reports/query.
Can some one share the thoughts on it.