Unified Cache for Translation Table Entries and Updates

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing data storage systems face challenges in efficiently managing metadata pages, particularly in updating and storing them in a way that allows for dynamic translation of logical to physical addresses, while minimizing storage overhead and maintaining data integrity.

Innovation Solution

The system employs a log-structured approach for metadata pages, where updates are applied to generate new versions of metadata pages, and these updates are stored in a translation table log. This approach allows for the dynamic updating of physical storage locations for metadata pages without overwriting existing versions, enabling efficient management of logical to physical address mappings.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If a traditional cache structure is used for storing translation table entries and updates, then access speed is improved, but storage overhead increases and data integrity becomes difficult to maintain

Engineering Contradiction:

Improveaccess speedVSAvoidstorage overhead

Core Design Contradiction:

SpeedVSQuantity of substance

Solution Approach 1:

The patent merges the translation table cache and update logs into a single unified cache structure. This consolidation eliminates redundant storage structures, reducing overall storage overhead while maintaining fast access capabilities. The unified cache stores both current translation table entries and their updates in one location, avoiding the need for separate cache and log structures.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The unified cache serves multiple functions simultaneously: it acts as a translation table cache for fast address translation, stores update logs for data integrity, and provides a single location for both reads and writes. This multi-functionality reduces the total storage required compared to having separate specialized structures for each function.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Reliability

If separate structures are used for translation table cache and update logs, then data integrity is improved, but device complexity increases

Engineering Contradiction:

Improvedata integrityVSAvoidstructure complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent combines the translation table cache and update log into a single unified cache structure, reducing device complexity by eliminating the need for managing separate structures. The unified structure uses a single hash table with entry types to distinguish between cache entries and log entries, simplifying the overall architecture while maintaining data integrity through type differentiation.

Inventive Principle:

Principle #5Merging (Combining)

3Quantity of substance

If metadata pages are updated in place, then storage space is conserved, but data integrity and version management become problematic

Engineering Contradiction:

Improvestorage spaceVSAvoiddata integrity

Core Design Contradiction:

Quantity of substanceVSReliability

Solution Approach 1:

The patent segments the unified cache into different entry types (cache entries and log entries) using a hash table structure. This segmentation allows the system to distinguish between current translation table entries and their updates, enabling proper version management and data integrity while efficiently utilizing storage space through a single consolidated structure.

Inventive Principle:

Principle #1Segmentation

4Quantity of substance

If a unified cache is used for both translation table entries and updates, then storage overhead is reduced, but access time may increase due to additional processing

Engineering Contradiction:

Improvestorage overheadVSAvoidaccess time

Core Design Contradiction:

Quantity of substanceVSLoss of time

Solution Approach 1:

The patent applies local quality by differentiating entry types within the unified cache using specific flags or fields in the hash table entries. This allows the system to quickly identify whether an entry is a cache entry or a log entry through localized markers, enabling efficient access patterns that minimize processing overhead despite the unified structure.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system performs preliminary classification of entries into cache and log types when they are inserted into the unified cache. This pre-categorization using entry types allows for optimized retrieval operations, as the system can directly access appropriate entries without requiring complex analysis during read operations, thus reducing access time.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12314186B1Techniques using a unified cache for storing and accessing translation table entries and updates

Publication Date: 2025.05.27 DELL PROD LP

AI summary

In at least one embodiment, processing can include: storing a metadata page at new physical location PA2 on non-volatile storage; and recording in a translation table (TT) log, a TT update to a mapping entry E1 of the TT, wherein the TT update updates E1 of the TT to map a logical address LA1 of the metadata page to PA2. Recording can include querying a unified hash table (UHT) in accordance with LA1 to determine whether the UHT includes a UHT entry mapping LA1 to a corresponding physical address or location of where the MD page is stored on non-volatile storage. Responsive to determining that the UHT includes a UHT entry for LA1, processing can include: determining whether the UHT entry is a cache type. If the UHT entry is the cache type, processing can include converting the UHT entry to a delta type entry of the TT log.