Key-Value Storage Metadata Mapping via Unified K2P Table
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage devices face inefficiencies in storing and managing key-value pair data and its associated metadata, leading to increased latency and memory requirements due to the need for additional mappings and separate storage locations.
Innovation Solution
A data storage device with a controller that generates a key-to-physical table mapping, where a single key is used to associate both the value and metadata, allowing for concurrent storage and retrieval, reducing the need for separate mappings and improving operational efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If metadata is stored separately from KV pair data with separate mappings, then metadata can be stored independently, but additional mappings are needed which increases latency and memory requirements
Solution Approach 1:
The patent merges the storage of metadata with the KV pair data by using the same key for both. The mapping structure includes a first slot for the value and a second slot for metadata, both accessed through the same key. This eliminates the need for separate mappings and reduces the number of memory references, thereby decreasing latency while maintaining metadata availability.
Solution Approach 2:
The key serves multiple functions: it maps to both the value and the metadata simultaneously. The mapping structure is designed to handle both the KV pair data and its associated metadata through a single key lookup, making the system more efficient and reducing the overhead of maintaining separate mapping tables.
2Adaptability or versatility
If separate mappings are used for metadata, then metadata can be addressed independently, but additional memory is required to store the mappings
Solution Approach 1:
The patent combines the mapping functions for both the value and metadata into a single mapping structure. Instead of requiring separate mapping tables for KV pair data and metadata, the system uses one unified mapping where the key points to both the value (first slot) and metadata (second slot), thereby reducing the total memory required for mappings.
Solution Approach 2:
The mapping structure is designed to be universal, handling both value retrieval and metadata access through the same key. This multi-functional approach eliminates the need for dedicated mapping memory for metadata, as the same mapping infrastructure serves both purposes.
3Adaptability or versatility
If metadata is transferred non-concurrently with KV pair data, then transfer flexibility is improved, but additional mappings are needed which increases complexity
Solution Approach 1:
The patent performs preliminary action by pre-allocating the second slot in the mapping structure for metadata before the metadata is actually transferred. When metadata arrives later, it can be directly placed into the pre-reserved second slot without requiring dynamic mapping creation or restructuring, thus maintaining flexibility while avoiding additional mapping complexity.
Data Source
AI summary
A data storage device includes a memory device and a controller to the memory device. The controller is configured to receive key value (KV) pair data having a key and a value from a host device and generate a mapping in a key-to-physical (K2P) table corresponding to the received KV pair data. The mapping includes a first slot for storing a physical address corresponding to the value and a second slot for storing a physical address corresponding to metadata associated with the KV pair data. When the associated metadata is sent to the data storage device, which may be non-concurrent to transferring the KV pair data, the mapping of the associated metadata is linked to a same key as the mapping of the KV pair data. Thus, using the mapping, the key of the KV pair data is associated with the KV pair data and the associated metadata.