Software RAID Crash-Dump Telemetry Storage for Server Resilience
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing RAID data protection solutions for information handling systems are limited by high costs, compatibility issues with various storage devices and CPUs, and inability to configure logical storage devices during pre-boot or initialization operations, particularly in server devices.
Innovation Solution
A software RAID engine is implemented in the operating system to provide a software Redundant Array of Independent Disk (RAID) system that configures and manages RAID infrastructure using direct-attached storage devices, allowing for RAID driver telemetry data storage during crash dump modes, and supports various storage devices and CPUs without dedicated hardware.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If hardware-based RAID data protection solution is used, then data protection capability is improved, but system cost increases significantly
Solution Approach 1:
The patent replaces hardware-based RAID mechanisms with a software-based RAID implementation. The software RAID engine runs on the host CPU and provides RAID functionality through software routines rather than dedicated hardware controllers, thereby eliminating the need for expensive hardware RAID cards while maintaining data protection capabilities through software-managed redundancy across multiple storage devices.
Solution Approach 2:
The patent creates software copies of RAID functionality that normally would require dedicated hardware. By implementing RAID logic in software form that can run on general-purpose processors, the system replicates the data protection functions of hardware RAID without requiring specialized hardware components, thus reducing system cost while preserving reliability.
2Reliability
If VROC data protection solution is used, then data protection is provided, but compatibility is limited to specific CPUs and storage devices
Solution Approach 1:
The patent implements a universal software RAID engine that can operate with any CPU architecture and any storage device type through standardized interfaces. The software RAID solution uses generic storage drivers and communication protocols that are platform-independent, allowing the same RAID implementation to function across different CPU families (Intel, AMD, ARM) and storage device types (HDD, SSD, NVMe, SAS, SATA) without requiring vendor-specific hardware or proprietary interfaces.
Solution Approach 2:
The patent introduces software abstraction layers and standardized interfaces that mediate between the RAID engine and underlying storage devices. These intermediary software components translate between different storage device protocols and the unified RAID management interface, enabling broad compatibility across diverse hardware platforms without requiring platform-specific RAID implementations.
3Productivity
If storage spaces data protection solution is used, then data aggregation is achieved, but configuration during pre-boot operations is not possible
Solution Approach 1:
The patent enables RAID configuration to be performed during pre-boot and initialization operations before the operating system fully loads. The software RAID engine includes initialization routines that can be executed during system boot, allowing RAID volumes to be configured and created in advance of normal OS operation, thereby enabling data aggregation functionality to be established during the initialization phase rather than requiring post-boot configuration.
4Ease of manufacture
If software RAID solution is implemented, then system cost is reduced, but ability to preserve RAID data during system crash is compromised
Solution Approach 1:
The patent implements a crash dump mechanism that reserves dedicated storage space in advance within the RAID logical storage subsystem. This pre-allocated crash dump area is prepared during RAID volume initialization and is specifically designated for storing RAID driver telemetry data and system state information in the event of a crash, providing beforehand protection against data loss during system failures.
Solution Approach 2:
The patent converts the potential harm of system crashes into a beneficial diagnostic opportunity by capturing and preserving RAID driver telemetry data during crash events. The crash dump functionality transforms the harmful system failure state into a useful source of diagnostic information that can help identify and resolve the underlying crash cause, thereby turning a negative event into a valuable debugging resource.
Data Source
AI summary
A crash-based RAID data storage system includes a chassis housing storage subsystems coupled to an operating system and a software RAID subsystem. During a runtime mode of the operating system, the software RAID subsystem reserves a RAID driver telemetry storage portion of a RAID logical storage subsystem provided by at least one storage device in the storage subsystems. In response to the operating system entering a crash dump mode from the runtime mode, the software RAID subsystem initializes respective controller devices in each of the storage subsystems, and during the crash dump mode of the operating system the software RAID subsystem stores RAID driver telemetry data generated by the software RAID subsystem in the RAID driver telemetry storage portion of the RAID logical storage subsystem provided by the at least one storage device using the respective controller device for that storage device.