Application-Based Checkpointing Control for Persistent Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems are susceptible to malware attacks and performance degradation due to resource-intensive backups, and they lack effective application-based checkpointing control, which can lead to data corruption and loss.
Innovation Solution
Implementing a file checkpoint manager with application-triggered checkpointing control, using a solid-state drive (SSD) with internal checkpointing technology, and configurable policies to manage and secure data on a per-file basis, inhibiting denial-of-service attacks and enabling secure data recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If operating system checkpointing is implemented to allow data restoration, then data recovery capability is improved, but the system becomes susceptible to malware attacks that can corrupt checkpoint data
Solution Approach 1:
The patent divides the checkpointing functionality into separate components: application-level checkpoint initiation and storage-device-level checkpoint management. This segmentation isolates the checkpoint data from operating system vulnerabilities, allowing applications to control checkpoint creation independently of OS security weaknesses.
Solution Approach 2:
The storage device acts as an intermediary between the application and the checkpoint data. The application sends checkpoint commands to the storage device, which then manages the checkpoint data independently. This intermediary layer prevents direct access to checkpoint data by malicious software, protecting against corruption while maintaining recovery capability.
2Reliability
If system backups are run on the host system to create comprehensive data copies, then data protection coverage is improved, but host system performance degrades due to significant resource consumption
Solution Approach 1:
The storage device performs checkpointing operations autonomously without requiring host system resources. The storage device's internal controller handles data copying and management independently, allowing the host system to maintain full productivity while the storage device self-manages backup operations.
Solution Approach 2:
The checkpointing function is extracted from the host system and relocated to the storage device. This extraction removes the performance burden from the host system, as the storage device independently handles all checkpointing operations including data copying and version management.
3Measurement precision
If application-based checkpointing control is implemented on a per-file basis, then data recovery precision is improved, but system complexity increases due to additional control mechanisms
Solution Approach 1:
The checkpointing system dynamically adapts to application needs through simple commands. Applications can issue checkpoint requests for specific files or directories based on their requirements, and the storage device automatically manages the complexity of per-file tracking and version control without requiring complex application-side implementation.
Solution Approach 2:
The storage device's checkpoint manager serves as an intermediary that handles the complexity of per-file checkpointing. Applications simply issue high-level commands to the storage device, which then manages the detailed tracking and control of individual file checkpoints, reducing application complexity while maintaining precision.
Data Source
AI summary
An embodiment of a semiconductor apparatus may include technology to receive an application-related checkpoint request corresponding to a file of a file system stored on a persistent storage media, and determine one or more checkpoint operations internal to the persistent storage media to perform the application-related checkpoint request. Other embodiments are disclosed and claimed.