Storage Controller Reconfiguration for SEU Logic Failure Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Solid State Disk (SSD) storage systems face logic errors due to Single Event Upsets (SEUs) caused by alpha-ions or electromagnetic radiation, leading to control logic failures, necessitating methods for reconfiguring storage controllers to maintain system integrity.
Innovation Solution
A method and apparatus for reconfiguring a storage controller when control logic fails, involving a processing unit that suspends access, stores current execution statuses, and uses reconfiguration control logic to reprogram the whole reconfigurable region of the storage controller, allowing for rapid recovery and resumption of operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the storage controller uses a reconfigurable region to recover from control logic failures, then the reliability is improved, but the device complexity increases
Solution Approach 1:
The storage controller is divided into a fixed region and a reconfigurable region. The fixed region contains stable control logic that cannot be modified, while the reconfigurable region can be dynamically reprogrammed to replace failed logic. This segmentation allows the system to maintain reliability through the reconfigurable portion without compromising the stability of the fixed portion, while managing complexity by isolating the reconfigurable elements to a specific region.
Solution Approach 2:
The reconfigurable region is pre-configured with backup control logic before failure occurs. When a failure is detected in the fixed region, the system can immediately switch to the pre-prepared reconfigurable logic without needing to design and implement a solution from scratch. This preliminary preparation of alternative logic paths enables rapid recovery and maintains reliability while avoiding the complexity of real-time logic synthesis.
2Reliability
If the processing unit suspends access to the storage controller when failure is detected, then the reliability is improved, but the loss of time increases
Solution Approach 1:
The system maintains execution status information in a ready state before failure occurs. When a failure is detected, the suspended access is minimal because the system has already prepared the necessary status data and can quickly resume operations after reconfiguration. This preliminary preparation of execution status reduces the time penalty associated with suspending and resuming access.
Solution Approach 2:
The system performs a rapid failover by quickly switching from the failed fixed region to the reconfigurable region without prolonged suspension. The suspension of access is kept brief, and the system rushes through the reconfiguration process by utilizing pre-prepared logic and status information, thereby minimizing the loss of time while maintaining reliability through the switch to functional logic.
3Reliability
If the reconfiguration control logic reprograms the whole reconfigurable region, then the reliability is improved, but the productivity decreases
Solution Approach 1:
The reconfigurable region is designed as a self-contained segment that can be reprogrammed independently of the fixed region. This segmentation allows the system to reprogram only the necessary reconfigurable logic without affecting the entire storage controller, thereby limiting the productivity impact to a specific portion while ensuring reliability through complete reconfiguration of the affected segment.
Solution Approach 2:
The reconfigurable region contains copied or replicated control logic that can be quickly loaded and activated. Instead of synthesizing new logic from scratch, the system uses pre-existing logic templates or copies that can be rapidly programmed into the reconfigurable region. This copying approach maintains reliability by providing verified working logic while reducing the time and resource cost of reconfiguration, thus minimizing the impact on productivity.
Data Source
AI summary
A method for processing return entities associated with multiple requests in a single ISR (Interrupt Service Routine) thread, performed by one core of a processing unit of a host device, is introduced. Entities are removed from a queue, which are associated with commands issued to a storage device, and the removed entities are processed until a condition is satisfied.