Sparse Cache for SMI-S Deletion Indication Translation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The Storage Management Initiative Specification (SMI-S) standard requires deletion indication filters that include all property values, but existing frameworks often lack necessary information, making it inefficient to redesign or implement full caching solutions due to high memory and CPU usage, especially with large numbers of instances.
Innovation Solution
A sparse cache is maintained by the SMI-S API layer to store previous attribute values, which are used to translate and send complete deletion indications to clients, avoiding the need for redesign and reducing performance inefficiencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If a full cache containing all possibly needed information is implemented, then deletion indication completeness is improved, but memory usage and CPU usage increase significantly
Solution Approach 1:
The patent implements a selective caching strategy where only specific property values that are frequently needed in deletion indications are cached, rather than caching all possible properties. This local quality approach caches only the essential attributes (such as Type, Capacity, OperationalStatus) that are commonly required, thereby reducing memory consumption while still providing complete deletion indications for most cases.
Solution Approach 2:
The patent uses partial action by implementing a sparse cache that contains only a subset of all possible property values. Instead of maintaining a complete cache of all properties for all instances, the system caches only the most critical properties needed for deletion indications, accepting that some less common properties may need to be retrieved through alternative means when required.
2Loss of information
If a full cache containing all possibly needed information is implemented, then deletion indication completeness is improved, but CPU usage increases due to polling overhead
Solution Approach 1:
The patent implements periodic polling of the sparse cache to detect changes in cached property values. The polling mechanism periodically checks for modifications to the cached data and triggers appropriate actions when changes are detected, rather than continuously monitoring all properties. This periodic action reduces CPU overhead compared to continuous full-cache monitoring while still maintaining deletion indication completeness.
Solution Approach 2:
The sparse cache is designed to be self-sufficient for handling common deletion indication scenarios. The cached property values are maintained and updated by the system itself through normal operation, reducing the need for external polling and processing. The cache automatically provides the necessary information for deletion indications without requiring extensive CPU intervention to retrieve data from underlying sources.
3Loss of information
If the framework is redesigned to provide all properties in underlying indication, then deletion indication completeness is improved, but implementation effort and risk of regressions increase
Solution Approach 1:
The patent introduces a sparse cache as an intermediary layer between the existing framework and the deletion indication mechanism. This intermediary cache stores essential property values that bridge the gap between the limited information provided by the framework and the complete information required by deletion indications. By adding this intermediate layer, the system achieves comprehensive deletion indications without requiring complex redesign of the underlying framework.
Solution Approach 2:
The patent creates a simplified copy of the necessary property information in the sparse cache, rather than requiring the entire framework to provide all properties. The cache contains copies of critical property values (Type, Capacity, OperationalStatus) that are sufficient for most deletion indication scenarios, eliminating the need for comprehensive framework redesign while maintaining indication completeness.
Data Source
AI summary
The subject disclosure is generally directed towards caching property values in a sparse cache for use in translating notifications to contain property values related to a source instance, e.g., for use in SMI-S compliant notifications (deletion indications). When a deletion indication translation needs properties that are unavailable in the current source instance, a cache is accessed to obtain the previous related property values. The deletion indication is translated based upon the related property values, and output, e.g., as a translated deletion indication to a client subscriber.