Dynamic Cache Eviction for Storage Capacity Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cache nodes face storage space issues due to excessive data accumulation, leading to random deletion of data without knowing the age of the deleted items, potentially removing valuable data while retaining unnecessary entries, resulting in system performance degradation.
Innovation Solution
A cache manager dynamically evicts cache entries by monitoring capacity levels, configuring time steps based on entry timestamps, and removing entries in increments until a predetermined reduction level is achieved, ensuring efficient use of storage space without harming system performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is continuously added to the cache node beyond its capacity, then the cache node stores more data for faster retrieval, but the cache node runs out of storage space and cannot accommodate new entries
Solution Approach 1:
The patent implements dynamic eviction by identifying and removing expired or least recently used cache entries when the cache node reaches its storage capacity. This allows the system to discard old data and make space for new data, maintaining optimal cache utilization without manual intervention. The eviction process recovers storage space by removing entries that are no longer needed, enabling continuous data addition to the cache.
2Volume of stationary object
If random deletion is used to free up storage space, then storage capacity is recovered quickly, but valuable data may be deleted while unnecessary entries are retained
Solution Approach 1:
The patent employs feedback mechanisms by continuously monitoring cache entry characteristics such as expiration timestamps and access patterns. The eviction decision is based on feedback from these monitored parameters, allowing the system to intelligently identify which entries should be removed. This feedback-driven approach ensures that only appropriate entries are evicted while preserving valuable data, avoiding the random deletion problem.
Solution Approach 2:
The system changes the parameter used for eviction decisions from random selection to based on expiration status and access patterns. By using these specific parameters, the cache manager can make informed decisions about which entries to remove, ensuring that expired or rarely accessed entries are evicted while retaining frequently accessed and non-expired entries.
3Volume of stationary object
If traditional eviction methods are used, then storage space is managed, but system performance degrades due to cache misses from premature data removal
Solution Approach 1:
The patent implements dynamic eviction policies that adapt to changing cache conditions rather than using fixed eviction rules. The system dynamically adjusts which entries to evict based on current cache state, access patterns, and expiration status. This dynamic approach optimizes the balance between storage space management and performance by making eviction decisions that minimize cache misses while maintaining adequate space for new entries.
Data Source
AI summary
A cache manager monitors storage capacity levels of a cache node storing data as cache entries. Upon determining a storage capacity threshold is reached, the cache manager begins evicting cache entries. In evicting the cache entries, the cache manager identifies a reduction level, and configures one or more time steps. The cache manager evicts cache entries within the one or more time steps until the reduction level is achieved.