Cache Server Recovery Control for Nonvolatile Memory Wear
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cache servers using nonvolatile memory face challenges in managing the wear-out of nonvolatile memory due to the finite number of program/erase cycles, leading to potential read errors and the need for effective wear suppression.
Innovation Solution
A cache server system that includes a processor to determine content availability, calculate delivery capability based on predicted read counts, and selectively recover content only when the delivery capability exceeds a reference value to manage wear on nonvolatile memory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If content is recovered and rewritten to nonvolatile memory after read errors, then content availability is improved, but nonvolatile memory wear increases
Solution Approach 1:
The system changes the parameter of recovery decision from binary (always recover or never recover) to conditional (recover only when delivery capability exceeds reference value). This parameter change resolves the contradiction by dynamically adjusting recovery behavior based on real-time delivery capability assessment, thereby improving content availability while controlling memory wear.
Solution Approach 2:
The system implements feedback by continuously monitoring delivery capability and using it to control the content recovery process. When delivery capability exceeds the reference value, recovery is permitted; otherwise, it is suppressed. This feedback mechanism resolves the contradiction by creating a self-regulating system that balances content availability with memory wear control.
2Reliability
If content is recovered and rewritten to nonvolatile memory, then read errors are suppressed, but program/erase cycle consumption increases
Solution Approach 1:
The system applies partial action by recovering content only when necessary (when delivery capability exceeds reference value) rather than always recovering. This partial recovery approach suppresses read errors in critical cases while consuming fewer program/erase cycles, thereby resolving the contradiction between reliability improvement and resource conservation.
3Productivity
If delivery capability is increased by recovering more content, then content availability improves, but nonvolatile memory wear accelerates
Solution Approach 1:
The system applies preliminary anti-action by setting a reference value threshold that prevents excessive recovery operations before they can cause significant memory wear. This preemptive constraint ensures that delivery capability enhancement does not accelerate memory wear, resolving the contradiction between productivity improvement and system longevity.
Data Source
AI summary
According to one embodiment, in a case where an error has occurred in reading a first content, a processor of a cache server calculates a delivery capability of the cache server. In a case where it is determined that the calculated delivery capability exceeds a first reference value, the processor does not execute recovering the first content. In a case where it is determined that the calculated delivery capability does not exceed the first reference value, the processor executes recovering the first content, delivering the recovered first content to a client via a network, and writing the recovered first content to a nonvolatile memory.


