Cache Eviction Using Timestamp Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Time threshold-based cache eviction methods often incorrectly identify frequently used entries for eviction, leading to suboptimal cache performance due to limited eviction processing, where less frequently used entries may remain in the cache instead of more frequently accessed ones.
Innovation Solution
Adjusting the timestamps of cache entries using an adjustment factor to extend the time-to-live (TTL) for frequently and recently used entries, thereby preventing them from being candidates for eviction and ensuring that less accessed entries are evicted instead, employing a combination of recency and frequency-based methodologies like LRU, LFU, and ARC.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If time threshold-based eviction is used, then eviction simplicity is improved, but frequently used entries are incorrectly evicted
Solution Approach 1:
The patent modifies the time threshold parameter dynamically based on entry characteristics. Instead of using a fixed time threshold for all entries, the system adjusts the effective threshold by adding a bias value derived from access frequency and recency metrics. This allows frequently accessed entries to effectively extend their TTL beyond the base threshold, preventing their incorrect eviction while maintaining the simplicity of threshold-based eviction.
2Productivity
If eviction limit is applied, then processing overhead is reduced, but frequently used entries remain in cache due to insufficient processing
Solution Approach 1:
The patent applies different treatment to different entries within the cache based on their access patterns. By calculating access frequency and recency metrics for each entry, the system identifies which entries deserve protection from eviction. This local differentiation allows the eviction limit to be applied more effectively, as the most important entries are identified and protected while still maintaining processing efficiency through selective evaluation.
3Reliability
If cache size is increased, then frequently used entries can be retained, but memory usage increases
Solution Approach 1:
The patent effectively increases the cache capacity for high-value entries by extending their TTL through parameter adjustment. Rather than physically expanding the cache, the system modifies the time parameter to allow frequently accessed entries to remain longer. This creates an effective capacity expansion where the same physical cache space serves more important data for longer periods, reducing the need for additional memory.
Data Source
AI summary
One example method includes a cache eviction operation. Entries in a cache are maintained in an entry list that includes a recent list and a frequent list. When an eviction operation is initiated or triggered, timestamps of last access for the entries are adjusted by corresponding adjustment values. Candidates for eviction are identified based on the adjusted timestamps of last access. At least some of the candidates are evicted from the cache.