Dynamic Garbage Collection for Data Storage Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data storage devices experience increased latency during periods of heavy workload due to sudden bursts of host commands, which further exacerbate latency issues during garbage collection operations, impacting the quality of service.
Innovation Solution
Implementing a dynamic garbage collection mechanism that disables GC operations when host command latency exceeds a threshold, prioritizing new data storage over data relocation by using available buffers instead of host blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If garbage collection operations are performed to relocate stored host data, then memory blocks are freed for new data storage, but host command latency increases and quality of service deteriorates
Solution Approach 1:
The patent implements dynamic garbage collection by adjusting GC operation frequency and intensity based on real-time workload conditions. When heavy workload is detected (high command latency), GC operations are throttled or paused. When workload is light, GC operations proceed normally to free memory blocks. This dynamic adjustment resolves the contradiction by making GC behavior adaptive rather than static.
Solution Approach 2:
The system changes operational parameters of garbage collection based on system state. Specifically, it monitors host command latency and adjusts GC parameters (such as GC threshold, GC interval, or GC priority) accordingly. When latency exceeds a threshold, GC parameters are modified to reduce GC activity, thereby resolving the trade-off between freeing memory blocks and maintaining low latency.
2Productivity
If the data storage device handles bursts of host commands, then productivity increases, but command processing latency increases
Solution Approach 1:
The patent applies dynamic workload management by monitoring incoming command rates and adjusting system behavior accordingly. During burst periods, the system dynamically adjusts GC operations and memory management to prioritize quick command completion over optimal memory utilization. This resolves the contradiction by allowing high throughput during bursts while accepting temporary latency increases.
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring command latency and workload intensity. Based on this feedback, it adjusts GC operation timing and priority. When high throughput is detected, the system feedback-loop adjusts GC to be less intrusive, thereby maintaining productivity while managing latency through real-time adaptation.
Data Source
AI summary
Devices, methods and apparatuses for dynamically enabling and disabling garbage collection operations during periods of burst in host command activity. Host command latency is monitored during operation of a data storage device. During normal operation, host data is stored in host blocks in a memory of the data storage device. During periods of burst, the host command latency may increase above a garbage collection disable threshold. In response, an electronic processor disables garbage collection operations. Additionally, during this time, host data is stored in buffers in the memory of the data storage device. When the host command latency decreases below a garbage collection enable threshold, the data storage device returns to normal operation and enables garbage collection operations.