Data Storage Reference State Management for Controller Load Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage systems face performance degradation and increased resource utilization when tracking logical data volumes due to the need for frequent updates of reference counts and metadata across multiple physical zones, especially during data deletions or snapshots, leading to inefficient memory accesses and processing demands.
Innovation Solution
The introduction of an intermediate page reference state and a 'recently freed list' allows the data storage controller to efficiently track logical data addresses with minimal reference count involvement, optimizing performance by delaying updates to reference states until necessary, thereby reducing the load on the storage controller and improving resource utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If reference counts and metadata are frequently updated across multiple physical zones to track logical data volumes, then data integrity and tracking accuracy are improved, but storage controller performance and resource utilization deteriorate
Solution Approach 1:
The patent applies preliminary action by pre-allocating zones and establishing reference state tracking structures before data operations occur. When a logical volume is created, the system pre-identifies physical zones and sets up reference count mechanisms, so that subsequent data operations can proceed without frequent metadata updates. This advance preparation reduces the overhead during actual data operations while maintaining tracking accuracy.
Solution Approach 2:
The patent introduces an intermediary mechanism in the form of a reference state indicator that mediates between the logical volume tracking requirement and the physical zone management. This intermediary layer allows the system to track logical data volumes without requiring frequent direct updates to all underlying physical zone metadata, thus reducing controller overhead while maintaining tracking precision.
2Reliability
If reference states are updated synchronously for each data operation, then data consistency is improved, but memory access time and processing delay increase
Solution Approach 1:
The patent implements periodic action by updating reference states asynchronously rather than synchronously with each data operation. The reference state indicators are updated at periodic intervals or batched together, allowing data operations to proceed without waiting for reference state updates. This periodic update mechanism maintains data consistency while significantly reducing memory access time and processing delays.
Solution Approach 2:
The patent ensures continuity of useful action by allowing data operations to proceed continuously without being blocked by reference state updates. The asynchronous update mechanism enables data writes and reads to occur without interruption, while reference state tracking continues in the background, thereby eliminating idle time and maintaining high throughput.
3Manufacturing precision
If all zones are allocated for new data with correct reference states, then allocation accuracy is improved, but controller processing load and complexity increase
Solution Approach 1:
The patent applies partial action by updating reference states for only some zones rather than all zones. When allocating new data, the system updates reference states selectively based on actual usage patterns and only when necessary, rather than performing comprehensive updates across all zones. This partial update approach maintains sufficient allocation accuracy while significantly reducing controller processing load.
Solution Approach 2:
The patent utilizes parameter changes by transitioning reference state indicators between different states (e.g., from incorrect to correct, or from one reference count to another) only when necessary. This selective parameter updating allows the system to maintain allocation accuracy for actively used zones while avoiding unnecessary processing for inactive or unchanged zones, thereby reducing overall controller complexity and load.
Data Source
AI summary
A data storage system can connect a data storage controller to a host and a data storage device. A first reference state corresponding to a first zone of the data storage device can be incremented in response to a first version of data being assigned to the first zone by the data storage controller. A second version of the data may be written to a second zone of the data storage device prior to populating a recently freed list with the first zone having an incorrect reference state. The first zone can be allocated by the data storage controller for new data without altering the incorrect reference state that is subsequently written to the first zone as directed by the data storage controller.