Client-Side Agent for Crash Consistent PCPI Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems lack the capability to generate crash consistent persistent consistency point images (PCPI) sets across multiple volumes or storage systems, leading to inconsistent data recovery due to asynchronous write operations, which is not supported by all file systems like Linux.
Innovation Solution
A client-side agent is implemented to manage data containers by obtaining read/write statistics, generating PCPIs, and ensuring no modifying operations occurred during image creation, with retries and error handling to achieve a crash consistent state.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If PCPI images are generated across multiple volumes or storage systems, then data recovery capability is improved, but data consistency deteriorates due to asynchronous write operations
Solution Approach 1:
The system performs preliminary actions by capturing read/write statistics before generating PCPI images and again after generation. This allows the system to verify that no modifying operations occurred during the image creation process, ensuring data consistency across multiple volumes while maintaining the ability to recover data from the generated images.
2Manufacturing precision
If crash consistent PCPI sets are generated, then data integrity is improved, but system complexity increases due to statistics verification and retry mechanisms
Solution Approach 1:
The system implements feedback by comparing read/write statistics captured before and after PCPI generation. This feedback mechanism automatically verifies data integrity and determines whether the generated PCPI set is crash consistent, eliminating the need for complex manual verification procedures while ensuring data integrity.
Solution Approach 2:
The system performs self-service by automatically verifying its own work through statistical comparison. The agent independently checks whether modifying operations occurred during PCPI generation and can retry the generation process if inconsistencies are detected, reducing the need for external intervention and complex verification systems.
3Productivity
If PCPI generation is performed without verifying modifying operations, then operation speed is improved, but data consistency deteriorates
Solution Approach 1:
The system captures read/write statistics as a preliminary action before PCPI generation begins. This allows the system to later verify data consistency by comparing pre and post generation statistics, ensuring that no modifying operations occurred during the process while maintaining efficient operation speeds through automated verification.
Data Source
AI summary
A technique for generating a crash consistent set of persistent consistency point image set associated with a storage space is provided. A client side agent obtains a first set of read/write statistics before causing storage systems to generate persistent consistency point images of volumes storing logical unit numbers comprising the storage space. Once the persistent consistency point images have been generated, the agent obtains a second set of read/write statistics. The first and second set of statistics are compared to determining if they are equal, thereby signifying that the set of persistent consistency point images is crash consistent.