Using query statistics
The warehouse manager has a number of areas of responsibility. One of these areas is maintaining the responsiveness of the system even when the query profiles change over time. This can be achieved by gathering statistics on the queries being physically executed against the data warehouse, and deducing the appropriate summaries to speed up common queries. Query statistics can be collected by the query manager as it intercepts any query hitting the database. In fact, some relational databases provide facilities to do this as a matter of course. If this information is available, the warehouse manager can be extended to various functions. The query manager is the system component that performs all the operations necessary to support the query management process. This system is typically constructed using a combination of user access tools, specialist data warehousing monitoring tools, native database facilities, bespoke coding, C programs, and shell scripts.
As with the load manager, the size and complexity of the query manager will vary between specific solutions. Unlike the load manager, the complexity of the query manager is driven by the extent to which the facilities are provided by user access tools or native database facilities. Practically, all the query manager is built in later development phases. Typically, the query manager is designed in the first build phase, once the database and user access tool technologies have been determined.