Storage Controller Handling for NVM Thermal Shutdown Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Thermal shutdown of non-volatile memory in data storage devices leads to degradation of quality of service (QoS) due to discarded host commands and halted operations, requiring improved thermal management schemes.
Innovation Solution
Implementing a thermal shutdown manager to optimize operations before, during, and after thermal shutdown by queueing host commands, performing background operations, and handling residual data, allowing the system to continue processing non-volatile memory-independent tasks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If thermal shutdown is implemented to protect non-volatile memory from overheating, then reliability is improved, but quality of service deteriorates due to discarded host commands and halted operations
Solution Approach 1:
The system performs preliminary actions by queueing host commands before thermal shutdown occurs and preserving command metadata (LBA, transfer length) in advance. This allows the system to maintain operational state information even when the NVM enters thermal shutdown, enabling resumption of operations without discarding pending commands and thus reducing quality of service degradation.
Solution Approach 2:
The controller acts as an intermediary between the host and the thermal shutdown state. It receives host commands, queues them during thermal shutdown, and manages the transition back to operational state. This intermediary role allows the system to protect against thermal damage while minimizing the impact on ongoing operations by coordinating command execution with thermal conditions.
2Reliability
If the storage device waits for temperature to cool below predefined limit before resuming operations, then thermal damage is prevented, but loss of time increases due to extended shutdown duration
Solution Approach 1:
The system performs preliminary actions by queueing host commands before thermal shutdown occurs and preserving command metadata (LBA, transfer length) in advance. This allows the system to maintain operational state information even when the NVM enters thermal shutdown, enabling resumption of operations without discarding pending commands and thus reducing quality of service degradation.
Solution Approach 2:
The system dynamically adjusts its behavior based on thermal conditions. During thermal shutdown, it transitions to a state where it queues commands and monitors temperature. When temperature falls below the threshold, it dynamically resumes normal operations with queued commands. This dynamic approach optimizes the balance between thermal safety and operational continuity, reducing overall time loss.
3Object-generated harmful factors
If ongoing host commands are discarded during thermal shutdown, then harmful factors are reduced, but loss of information increases due to partial completion
Solution Approach 1:
The system performs preliminary actions by queueing host commands before thermal shutdown occurs and preserving command metadata (LBA, transfer length) in advance. This allows the system to maintain operational state information even when the NVM enters thermal shutdown, enabling resumption of operations without discarding pending commands and thus reducing quality of service degradation.
Solution Approach 2:
The system recovers command information that would otherwise be lost. By preserving command metadata (starting LBA, transfer length) in the controller's memory during thermal shutdown, it can recover and resume incomplete commands after thermal conditions are resolved, rather than discarding them entirely. This recovery mechanism minimizes information loss while still protecting against thermal damage.
Data Source
AI summary
Methods and apparatus for thermal management in data storage devices are provided. A data storage device includes a first non-volatile memory (NVM), and one or more processors coupled to the first NVM. The one or more processors are configured to receive, from a host, a host command to be performed in connection with the first NVM, the host command comprising a starting logical block address (LBA) and a transfer length, store the starting LBA and the transfer length in a second NVM, attempt to perform the host command, detect, before completion of the performance of the host command, a thermal shutdown state of the first NVM, and invalidate, using the stored starting LBA and the stored transfer length, residual data stored in the first NVM in association with partial completion of the host command prior to the thermal shutdown state.