Main Memory Partitioning for Parallel Database Query Processing

The patent segments the database system into multiple computing devices, each handling specific portions of data and queries. This distributed architecture divides the monolithic system into manageable units that can process operations concurrently, thereby increasing processing speed while maintaining controlled complexity through modular design.

The patent transitions from a single-dimensional centralized processing model to a multi-dimensional distributed architecture across multiple computing devices. This dimensional expansion allows parallel processing of queries and data operations, significantly improving processing speed by utilizing multiple processing dimensions simultaneously.

The patent implements continuous query processing across multiple computing devices where operations do not wait for sequential completion. Multiple queries are processed concurrently in parallel streams, eliminating idle time and ensuring continuous useful action throughout the system, thereby reducing overall query execution time.

The patent performs preliminary actions by pre-processing and caching data in memory buffers before queries are executed. This preliminary preparation of data in accessible memory formats reduces the time required for actual query execution, as data is already positioned and formatted for rapid retrieval and processing.

The patent segments centralized data storage into distributed storage across multiple computing devices. Each device maintains portions of the database in its local memory and storage, allowing independent access and processing. This segmentation enables linear scalability as new devices can be added to the distributed architecture without modifying the core storage structure.

The patent creates a universal distributed storage architecture where each computing device can serve multiple functions: storing data locally, processing queries against local data, and participating in distributed query execution. This multi-functionality enhances processing scalability as each device contributes to both storage capacity and processing power simultaneously.

The patent implements self-service memory management where each computing device autonomously manages its own allocated memory resources. Devices independently allocate and manage their memory buffers, cache structures, and data storage without requiring centralized memory control, thereby improving data access speed through local memory operations while keeping memory management complexity distributed and manageable.

The patent allocates excessive memory resources to each computing device beyond strict minimum requirements, allowing devices to pre-load and cache data proactively. This partial over-allocation of memory enables faster data access through local caching, while the distributed nature of memory management across multiple devices prevents any single device from becoming a memory management bottleneck.