Persistent Memory Indexing Maps for Fault Metadata Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Information handling systems (IHS) face challenges in tracking, synchronizing, and recovering metadata changes during hardware faults, particularly in persistent memory devices, where battery failures or weak statuses restrict write operations and lead to read-only modes, causing data loss and system instability.
Innovation Solution
A method and system that utilize a firmware interface to generate and synchronize indexing maps across multiple persistent memory devices, enabling write operations and coordinating changes during hardware faults by identifying writeable maps and distributing delta metadata updates, ensuring metadata integrity and system resilience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If NVDIMM battery status is weak or failed, then data preservation is ensured by placing memory regions in read-only mode, but write operations are blocked causing metadata updates to fail
Solution Approach 1:
The patent segments metadata storage into two separate indexing maps: a first indexing map stored on the persistent memory device (NVDIMM) and a second indexing map stored on a different storage device. This segmentation allows the system to bypass the read-only restriction on the NVDIMM by writing metadata updates to the second indexing map instead, while still maintaining data integrity through the protective read-only mode on the original device.
Solution Approach 2:
The patent introduces a intermediary mechanism (the second indexing map on a separate storage device) that acts as a mediator between the firmware and the protected NVDIMM. When the NVDIMM is in read-only mode due to battery issues, the intermediary allows metadata updates to be performed externally, and then selectively applied to the original device when conditions permit, thus resolving the contradiction between protection and updatability.
2Reliability
If metadata is stored only on the persistent memory device, then data integrity is maintained, but metadata loss occurs when hardware faults prevent write operations
Solution Approach 1:
The patent applies local quality by creating different copies of the indexing map with different characteristics: the first indexing map maintains strict integrity by remaining on the protected NVDIMM in read-only mode, while the second indexing map is created with update-friendly properties by storing it on a separate storage device that allows write operations even when the NVDIMM battery is weak or failed.
Solution Approach 2:
The patent prepares a backup indexing map in advance on a separate storage device before hardware faults occur. This beforehand cushioning ensures that when the NVDIMM enters read-only mode due to battery issues, the metadata information is already preserved in the second indexing map, preventing metadata loss and enabling recovery operations.
3Reliability
If the system uses read-only mode for protected memory regions, then data protection is achieved, but system functionality is reduced due to inability to update metadata
Solution Approach 1:
The patent adds another dimension to the storage architecture by introducing a second indexing map on a separate storage device, creating a multi-dimensional metadata storage solution. This allows the system to maintain the protective read-only mode on the original NVDIMM while simultaneously enabling metadata updates through the additional dimension provided by the second indexing map.
Data Source
AI summary
A method and data processing device for enabling a write operation to track meta-data changes during a hardware fault in an information handling system (IHS). The method includes generating an indexing map to track memory space attributes of a persistent memory device. The method includes generating a subsequent indexing map that is a duplicate of a first indexing map. The method includes communicatively linking each of the indexing maps. The method includes distributing a subsequent indexing map to one or more memory devices. In response to detection of an update to meta-data associated during a hardware fault, the method includes identifying an indexing map that is stored on a writeable memory device. In response to detection of the hardware fault the method includes writing memory space attributes to the writeable indexing map. The method includes synchronizing a master indexing map to each other indexing map to coordinate changes to the memory space attributes.