Snapshot Deletion Request Batching in Storage Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage systems are inefficient in processing snapshot deletion requests, particularly for objects with replication or schedule features that generate frequent deletion requests, leading to suboptimal snapshot deletion performance and delayed release of storage space.
Innovation Solution
A method that determines objects with frequent snapshot deletion requests and adjusts the snapshot deletion request list to aggregate and prioritize these requests, using thresholds for the number and waiting time of requests to optimize the execution of snapshot deletion operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If snapshot deletion requests are processed individually in a first-come-first-served manner, then the system maintains simple processing logic, but the snapshot deletion performance is suboptimal and storage space is not released in time
Solution Approach 1:
The patent merges multiple snapshot deletion requests into batches based on object grouping and time window criteria. Requests for the same object are combined into a single deletion operation, and multiple batches are processed in parallel, transforming individual sequential processing into consolidated parallel execution to improve throughput
Solution Approach 2:
The system dynamically adjusts the snapshot deletion strategy by introducing a time window parameter and object-based grouping criteria. The batch size and processing frequency adapt based on the number of requests and object characteristics, allowing the system to optimize between individual processing and batch processing based on real-time conditions
2Productivity
If multiple snapshot deletion requests for the same object are aggregated together, then the deletion efficiency for those specific requests improves, but requests from various objects remain disordered and cannot be processed efficiently
Solution Approach 1:
The patent segments the snapshot deletion request list into multiple batches based on object grouping and time window criteria. Each batch contains deletion requests for specific objects within a defined time window, allowing organized parallel processing of otherwise disordered requests from various objects
Solution Approach 2:
The system introduces a new dimension of organization by adding time window grouping to the existing object-based grouping. Requests are first grouped by object, then further segmented by time windows, creating a two-dimensional organization structure that transforms disordered requests into systematically arranged batches
3Quantity of substance
If the storage system is configured with thousands of objects, then the system capacity increases, but snapshot deletion becomes a lengthy process and storage space cannot be released in time
Solution Approach 1:
The patent implements periodic action by processing snapshot deletion requests in time-based batches rather than continuously or all at once. The time window parameter divides the deletion process into periodic intervals, allowing the system to handle thousands of objects in manageable chunks while maintaining steady progress and releasing space continuously
Solution Approach 2:
The system dynamically processes deletion batches based on the volume of requests and object count. By introducing adjustable time windows and object-based grouping, the processing speed and batch size adapt to the system scale, enabling efficient handling of thousands of objects without excessive total processing time
Data Source
AI summary
Techniques perform snapshot deletion. Such techniques involve: determining an object associated with a to-be-executed snapshot deletion request in a snapshot deletion request list of a storage system; in response to the object being included in a predefined set of objects, determining information associated with the to-be-executed snapshot deletion request, the information including at least one of: a number of snapshot deletion requests in the snapshot deletion request list which correspond to snapshots associated with the object, and a waiting time of the to-be-executed snapshot deletion request in the snapshot deletion request list; determining, based on the information and from the snapshot deletion request list, a set of snapshot deletion requests associated with the object; and deleting snapshots corresponding to snapshot deletion requests in the set. Accordingly, the performance of snapshot deletion operations may be improved without any impact on other service on the storage system.