Indirect database queries with large olap cubes
AssigneeMicrosoft Corporation
InventorsCelestine Joseph, Joyce Behrendt, Joshua Chait Caplan
Life of this patent
Estimating the term…
Abstract
A large OLAP cube includes data and dimensions not required for a particular query, or for a collection of queries. A response to a query is generated by using the large OLAP cube rather than using an OLAP cube tailored to the query, and rather than querying the relational database directly. The large OLAP cube may be created after identifying a set of frequently sought data and/or filters used in queries to the relational database. Data may be staged by creating a smaller OLAP cube with data extracted from the large OLAP cube. The cubes may then be queried, browsed, and used to prepare reports or generate flat files.
Claims20
A method of facilitating querying data of a relational database, the method comprising the steps of:
- receiving a query whose processing would require access to only a relatively small set of data of the relational database, namely, a data set which includes less than twenty percent of at least one of the following aspects of the relational database:
- data records, storage space, tables, rows, columns, keys;
- obtaining a large online analytical processing cube, namely, an online analytical processing cube which includes a derived version of the relatively small set of data that is required for the query, and which also includes a derived version of some other data of the relational database which are not required for the query;
- and
- using the large online analytical processing cube to generate a response to the query rather than querying the relational database directly.
A computer-readable medium configured with data and instructions for performing a method for facilitating querying data of a relational database, the method comprising the steps of:
- receiving queries whose processing collectively would require access to only a relatively small set of data of the relational database, namely, a data set which includes less than ten percent of at least two of the following aspects of the relational database:
- data records, storage space, tables, rows, columns, keys;
- and
- using a large online analytical processing cube to generate responses to the queries rather than querying the relational database directly, the large online analytical processing cube including a derived version of the relatively small set of data that is required for the queries, and also including a derived version of other data of the relational database which are not required for any of the queries.
A computer system comprising:
- a logical processor;
- a relational database configuring memory in operable communication with the logical processor;
- a large online analytical processing cube configuring memory in operable communication with the logical processor;
- a smaller online analytical processing cube configuring memory in operable communication with the logical processor;
- and
- a query interface in operable communication with the logical processor, the query interface configured to receive a query and to then access the smaller cube if all data required for a query response are available in the smaller cube, otherwise to access the large cube if all data required for a query response are available in the large cube, and otherwise to access the relational database.
Dependencies are inferred from the claim text.