Methods and apparatus for tuple management in data processing system
US6751619B1 · kind B1 · utility
Assignee
Inventors
Key dates
| Filing date | Mar 15, 2000 |
| Grant date | Jun 15, 2004 |
| Priority date | — |
| Expiry date | Mar 15, 2020 |
Classification
- Technology area (CPC Y)Emerging Cross-Sectional Technologies
- CPC primaryY10S707/99933
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
A data processing system stores information in tuple space as tuples that are accessible by multiple entities. Methods, apparatus, computer-readable media, and data structures provide efficient extensions to tuple space coordination languages for example Linda, that increase concurrency by lessening tuple removal, without requiring compile time analysis, altering existing primitives, or adding new primitives. Traces are used to analyze tuple space access in distributed systems, resulting in optimizations based upon certain conditions which, if met, enable a tuple to remain visible in tuple space without blocking, so that other processes can continue to read the tuple while a first process is updating the tuple. A run-time optimization modifies the conditions if the execution is in a closed system that is known not to intentionally contain deadlock, further improving performance.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.