Page Buffer Pool Hashing for Storage Space Reclamation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional ring-based log architectures face challenges with large stepwise resource freeing, leading to unsmooth and unpredictable space reclaiming, causing IO write throttling and performance degradation due to sensitivity to flush latency.
Innovation Solution
A page buffer pool is generated within a data journal of a storage node, where each page buffer is hashed and stored in a page descriptor, allowing for immediate release and reclamation without dependencies, enhancing storage system performance by avoiding unsmooth space reclaiming.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional ring-based log architecture is used with sequential space reclamation, then data integrity is maintained through ordered flushing, but space reclamation becomes unsmooth and unpredictable causing IO write throttling
Solution Approach 1:
The patent segments the sequential flush dependency by introducing independent hash-based page buffer identification. Each page buffer is assigned a unique hash value that allows individual tracking and reclamation without waiting for sequential flush completion. This segmentation breaks the rigid sequential dependency while maintaining data integrity through hash verification.
Solution Approach 2:
The patent introduces hash values as an intermediary mechanism between page buffers and the ring structure. These hash values serve as mediators that enable indirect identification and reclamation of page buffers without requiring sequential access or dependency on the ring tail position, thus smoothing space reclamation while preserving data integrity.
2Stability of the object's composition
If ring tail dependency is enforced for space reclamation, then ordered data flushing is ensured, but IO latency increases due to waiting for flush completion
Solution Approach 1:
The patent applies preliminary action by pre-computing and storing hash values for each page buffer when it is allocated. This preliminary hashing enables immediate identification and reclamation of page buffers without waiting for flush operations to complete, reducing IO latency while maintaining flush ordering through the hash-based tracking mechanism.
Solution Approach 2:
The patent creates a hash-based copy or representation of page buffer identification that is independent of the physical ring structure. This hash copy allows the system to track and reclaim page buffers without being constrained by the ring tail position, thereby reducing IO latency while preserving the logical ordering of flush operations.
3Reliability
If page buffers are held until ring tail moves, then data consistency is maintained, but adaptiveness decreases due to rigid dependency
Solution Approach 1:
The patent introduces dynamics by replacing the static ring tail dependency with dynamic hash-based page buffer identification. Page buffers can be dynamically identified, tracked, and reclaimed based on their hash values rather than their position in the sequential ring structure. This dynamic approach maintains data consistency through hash verification while significantly improving system adaptiveness and responsiveness.
Data Source
AI summary
A method, computer program product, and computing system for generating a page buffer pool within a data journal of a storage node. A plurality of IO operations may be processed on a storage array using the storage node. A plurality of pages may be persisted in a plurality of page buffers within the page buffer pool based upon, at least in part, the processing of the plurality of IO operations. For each page buffer in the page buffer pool, a hash of the page buffer may be generated, thus defining a page buffer hash. The page buffer hash may be stored in a page descriptor associated with the page buffer.