Metablock Erase Scheduling for Low-Latency Write Accumulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data storage devices experience performance drops and memory block endurance issues due to the latency caused by the block exchange process, where a metablock is closed and a new one is opened for writing, with erase operations taking significant time and impacting concurrent operations.
Innovation Solution
A scheduling system that proactively erases target metablocks when the sum of upcoming write commands exceeds the available space in the currently operating metablock, allowing concurrent or parallel erase operations with write operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If a new metablock is erased before writing (pre-erasing), then the latency associated with block exchange is reduced, but memory block endurance is negatively impacted
Solution Approach 1:
The scheduling system performs preliminary erasure of target metablocks by identifying when the sum of upcoming write commands exceeds available space in the current metablock, then proactively initiating erase operations on target metablocks before they are needed. This preliminary action eliminates the 5ms+ latency that would otherwise occur when erasing at the moment of block exchange.
Solution Approach 2:
The scheduling system continuously monitors the command queue to determine the sum of sizes of upcoming write commands, compares this sum against available space in the current metablock, and uses this feedback to dynamically decide when to initiate proactive erasure operations. This feedback mechanism ensures erasure occurs at the optimal moment to balance latency reduction and endurance preservation.
2Reliability
If erase operations are performed during idle time, then concurrent operations during erase are avoided, but block exchange latency increases
Solution Approach 1:
Instead of waiting for idle time to perform erasure operations, the system takes preliminary action by proactively erasing target metablocks as soon as the command queue analysis indicates they will be needed. This eliminates the waiting period and ensures target blocks are ready immediately when required, preventing latency.
Solution Approach 2:
The system dynamically adjusts the timing of erasure operations based on real-time analysis of the command queue and upcoming write commands. Rather than using a static schedule or waiting for idle periods, the erasure timing is continuously adapted to match actual workload demands, allowing optimal concurrency between write and erase operations.
3Productivity
If the sum of upcoming write commands is monitored and compared to available space, then proactive erasure timing is optimized, but system complexity increases
Solution Approach 1:
The scheduling system performs multiple functions using the same command queue analysis mechanism: it identifies upcoming write commands, calculates their total size, compares against available space, and triggers erasure operations. This multi-functional approach avoids adding separate monitoring systems while achieving optimized erasure timing.
Solution Approach 2:
The scheduling system uses its existing command queue management infrastructure to self-determine when erasure operations are needed. By leveraging the same data structures and analysis routines already required for write command scheduling, the system achieves proactive erasure optimization without requiring additional complex monitoring or control mechanisms.
Data Source
AI summary
A data storage device includes a scheduling system that monitors a number of commands in a command queue and determines which of the commands in the command queue are write commands. The scheduling system also determines a sum of a size of the write commands in the command queue and compares it to an amount of available space in a currently operating metablock. If the sum of the size of the write commands exceeds the available amount of space in the currently operating metablock, the scheduling system identifies a target metablock and proactively initiates an erase operation on the target metablock prior to the currently operating metablock reaching capacity.