Data Warehouse – Data Warehouse Principle and NOSQL

By Nosql

Using MongoDB, CouchDB and related technologies, we can query faster, is this still valid?

“A copy of transaction data, specifically for query and analysis.” (R. Kimball The Data Warehouse Toolkit, 1996

I mean, are we really Need to reorganize our data into an OLAP solution for analysis for analysis? More specifically, can NoSQL (not necessarily OLAP modeling) be used to achieve drill-down, slicing and dice and other reports for analysis? Can you also overcome the “data subset” query limitation of OLAP and report the entire data world with NoSQL?

In my estimation, the OLAP subset or structure will not disappear, and may become more common for some reasons. There is no special order: f) Map-reduce is what you get in many cases. Mongodb is faster The aggregation pipeline is more stable; u) An important problem of NoSQL is the lack of connections or relationships. This means that your underlying data must be ugly to support many OLAP reports; b) It is worth building “discarded” or volatile data sub Set, just to keep a clean main table/collection; a) NoSQL is very suitable for redundant data sets: there is no need to create tables or even schemas, it is easy to rotate and kill the collection; r) NoSQL is easier to expand than other data sets, and Not SQL; d) fledgling startups can avoid the costs and resources required to support two database technologies (one for OLAP and the other for OLTP); and, b) use massage datasets and you will find you The back-end/front-end code is easier and easier to manage; and, c) the unparalleled speed advantage of the pre-made data set with its own pre-made index.

Using MongoDB, CouchDB and related technologies, We can query faster, will this still work?

“A copy of transaction data, specifically for query and analysis.” (R. Kimball The Data Warehouse Toolkit, 1996

I mean, are we really Need to reorganize our data into an OLAP solution for analysis for analysis? More specifically, can NoSQL (not necessarily OLAP modeling) be used to achieve drill-down, slicing and dice and other reports for analysis? Can it also overcome the “data subset” query limitation of OLAP and use NoSQL to report the entire data world?

In my estimation, the OLAP subset or structure is not Will disappear, and may become more common for some reasons. There is no particular order: f) Map-reduce is what you get in many cases. Mongodb is more stable through faster aggregation pipelines; u) An important part of NoSQL The problem is the lack of connections or relationships. This means that your underlying data must be ugly to support many OLAP reports; b) It is worth building “discarded” or volatile subsets of data, just to keep a clean main table/collection; a) NoSQL is very suitable for redundant data sets: there is no need to create tables or even schemas, it is easy to rotate and kill the collection; r) NoSQL is easier to scale than other data sets, not SQL; d) fledgling startups can avoid support The cost and resources required for the two database technologies (one for OLAP and the other for OLTP); and, b) using the massage data set, you will find your back-end/front-end code easier and easier to manage; And, c) the unparalleled speed advantage of the prefabricated data set with its own prefabricated index.

Leave a Comment

Your email address will not be published.