Host Anomaly-Based Snapshot Generation in Storage Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional storage systems face inefficiencies in data management and security, particularly in detecting and responding to security threats, and optimizing storage operations across multiple flash drives without redundant processes.
Innovation Solution
The implementation of a storage system architecture that offloads device management responsibilities from storage drives to controllers, utilizing non-volatile random access memory (NVRAM) for quick data buffering and direct-mapped flash storage systems that initiate and control processes at the operating system level, enabling improved latency and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If traditional storage systems manage data directly on storage drives, then data management is distributed, but device complexity and redundant processes increase
Solution Approach 1:
The patent extracts device management responsibilities from individual storage drives and consolidates them into a dedicated storage controller. The controller handles namespace management, data placement, and coordination across multiple drives, eliminating redundant management processes on each drive and reducing overall system complexity.
Solution Approach 2:
The storage controller serves as a universal management entity that handles multiple functions including data routing, namespace management, wear leveling, and security operations across all connected storage drives. This multi-functional approach replaces multiple specialized management functions distributed across drives with a single coordinated controller.
2Reliability
If snapshots are taken frequently for security monitoring, then threat detection capability improves, but storage space and system overhead increase
Solution Approach 1:
The patent implements selective snapshot creation based on anomaly detection. Instead of taking snapshots uniformly across all storage operations, the system monitors storage operations for anomalies and triggers snapshots only when suspicious patterns are detected, concentrating monitoring resources on critical events rather than all operations.
Solution Approach 2:
The system continuously monitors storage operations and uses feedback from anomaly detection algorithms to dynamically adjust snapshot creation. When anomalies are detected, the system increases snapshot frequency; when operations are normal, it reduces or eliminates snapshots, creating a responsive monitoring system that adapts to actual security needs.
3Reliability
If write operations are performed on every data change for consistency, then data integrity is maintained, but write amplification and performance decrease
Solution Approach 1:
The storage controller maintains a write buffer in volatile memory where write operations are staged before being committed to non-volatile storage. This preliminary buffering allows the system to acknowledge writes quickly to the host while batch-processing and optimizing the actual physical write operations to storage media, reducing write amplification.
Solution Approach 2:
The patent combines multiple small write operations into larger batch writes when possible. The controller accumulates write requests in a buffer and consolidates them into fewer, larger write operations to the storage media, maintaining data consistency while significantly reducing the total number of write operations and improving throughput.
Data Source
AI summary
An illustrative method includes a data protection system detecting a request provided by a host to perform an operation with respect to a storage system, detecting, based on the request, an anomaly associated with the host, and directing, based on the detecting the anomaly associated with the host, the storage system to generate a recovery dataset for data maintained by the storage system.