Exploitation of Correlation Between Original and Desired Data Sequences During Run Generation
AssigneeHewlett-Packard Development Company, L.P.
InventorsGoetz Graefe, Harumi Kuno
Life of this patent
Estimating the term…
Abstract
A computer executed method of exploiting correlations between original and desired data sequences during run generation comprises, with a processor, adding a number of data values from a data source to a first memory device, the first memory device defining a workspace, determining whether the data values within the workspace should be output in ascending or descending order for a number of runs, and writing a number of the data values as a run to a second memory device in the determined order.
Claims15
A computer executed method of exploiting correlations between original and desired data sequences during run generation comprising:
- with a processor:
- adding a number of data values from a data source to a first memory device, the first memory device defining a workspace;
- determining whether the data values within the workspace should be output in ascending or descending order for a number of runs;
- and
- writing a number of the data values as a run to a second memory device in the determined order.
A system for exploiting an incidental sort order of data during run generation comprising:
- a first memory device, the first memory device defining a workspace;
- a second memory device that stores generated runs;
- and
- a processor that:
- retrieves data values from a data source communicatively coupled to the system and adds the data values to the workspace;
- determines whether the data values within the workspace should be sorted in ascending or descending order;
- and
- writes a number of the data values as a run to the second memory device in the determined order.
A computer executed method of exploiting correlation between requested and on-disk data sequence during a disk scan operation comprising:
- with a processor:
- scanning a number of data values from a disk in response to a number of requests for data values;
- determining whether the order in which data values are being requested correlate to the order in which data values are being scanned from the disk;
- and
- determining whether to reverse the order of the disk scan based on the distribution of data values being requested.
Dependencies are inferred from the claim text.