Selective Cache Metadata Mirroring for Dual Controller Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In dual controller storage systems, the existing method of mirroring data at the granularity of cache pages wastes communication resources by transmitting the entire cache page even when only a small part of the data is updated, leading to performance bottlenecks.
Innovation Solution
The solution involves writing metadata to a cache and identifying the address range of the metadata, then copying only the data within that address range to a peer processor, rather than replicating the entire cache page, thereby reducing communication resources and improving system performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is mirrored at the granularity of cache page, then data synchronization between storage processors is ensured, but communication resources are wasted and system performance is bottlenecked
Solution Approach 1:
The patent segments the cache page into smaller granularities by introducing the concept of metadata blocks and data blocks. Instead of mirroring the entire cache page, only the specific metadata block that has changed is identified and mirrored to the peer storage processor. This segmentation allows selective synchronization of only the necessary data portions, reducing communication overhead while maintaining data consistency.
Solution Approach 2:
The patent extracts the metadata from the cache page structure and treats it as a separate mirroring target. By identifying the specific metadata block address range that has changed, the system extracts only this portion for synchronization rather than transferring the entire cache page. This extraction principle directly addresses the waste of communication resources by transmitting only the essential changed information.
2Reliability
If entire cache page is copied to peer storage processor, then data consistency is maintained, but data transmission volume and time increase
Solution Approach 1:
The patent divides the cache page into metadata blocks and data blocks, allowing independent identification and mirroring of only the changed metadata block. This segmentation enables the system to maintain data consistency by mirroring only the specific block that has changed, rather than copying the entire cache page, thereby significantly reducing data transmission time.
Solution Approach 2:
The patent applies partial action by mirroring only the necessary portion of data (the changed metadata block) rather than the entire cache page. This partial mirroring approach is sufficient to maintain data consistency across storage processors while minimizing the time and resources required for data transmission.
3Reliability
If whole cache page is transmitted for metadata update, then metadata synchronization is achieved, but communication interface workload increases
Solution Approach 1:
The patent segments the cache page structure into distinct metadata blocks and data blocks, enabling the identification and transmission of only the specific metadata block that has changed. This segmentation reduces the workload on the communication interface by eliminating the need to transmit unchanged portions of the cache page, thereby improving overall communication throughput.
Solution Approach 2:
The patent extracts the changed metadata block from the cache page and transmits only this extracted portion to the peer storage processor. This extraction approach achieves metadata synchronization while significantly reducing the communication interface workload compared to transmitting the entire cache page.
4Device complexity
If cache page level mirroring is used, then simplicity of implementation is maintained, but system performance is limited
Solution Approach 1:
The patent introduces segmentation of the cache page into metadata blocks and data blocks, adding a layer of complexity to the mirroring mechanism. However, this segmentation enables selective mirroring of only changed metadata blocks, which significantly improves storage system performance by reducing communication overhead and enabling more efficient use of available bandwidth.
Data Source
AI summary
Techniques perform data storage. Such techniques may involve writing metadata to a first cache of a first processor, the metadata indicating allocation of a storage resource to user data. Such techniques may further involve determining an address range of the metadata in the first cache. Such techniques may further involve copying only data stored in the address range in the first cache to a second cache of a second processor. Accordingly, the data transmission volume between two processors is reduced, which helps to improve the overall performance of a storage system.