Container Image Layer Storage Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Storing all layers of a container image in the same storage device ignores varying access rates, leading to unnecessary consumption of valuable local hard-disk space, as relatively unused layers occupy space alongside frequently accessed ones.
Innovation Solution
Determining a score for each layer based on usage and storing higher-scoring layers in high-performance storage destinations like SSDs or local hard drives, while lower-scoring layers are stored in lower-performance destinations such as remote cloud storage, optimizing storage usage and accessibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If all layers of a container image are stored in the same storage device, then storage simplicity is maintained, but valuable local hard-disk space is unnecessarily consumed by relatively unused layers
Solution Approach 1:
The patent segments the storage of container image layers by separating frequently accessed layers from less frequently accessed layers into different storage destinations. This segmentation allows the system to optimize storage space by placing only essential layers on local hard drives while storing less critical layers on remote cloud storage, thereby reducing local storage consumption without significantly complicating the storage management architecture.
Solution Approach 2:
The patent applies local quality by assigning different storage characteristics to different layers based on their access patterns. Frequently accessed layers are stored with high availability on local storage devices, while less frequently accessed layers are stored on remote cloud storage. This differential treatment optimizes the overall storage system by matching storage quality to actual usage requirements.
2Ease of operation
If all layers are stored together in the same storage device, then data accessibility is simplified, but frequently accessed layers cannot be quickly retrieved while less frequently used layers occupy valuable space
Solution Approach 1:
The patent segments layers into different storage destinations based on access frequency, placing frequently accessed layers on local storage devices for quick retrieval while storing less frequently accessed layers on remote cloud storage. This segmentation enables the system to achieve fast retrieval speeds for critical layers while maintaining a simplified access model through automated layer identification and placement.
Solution Approach 2:
The patent introduces an intermediary layer (the storage management system) that automatically identifies frequently accessed layers and places them on appropriate storage destinations. This intermediary handles the complexity of differential storage placement, allowing users to access layers through a simplified interface while the system optimizes retrieval speeds by pre-positioning frequently accessed layers on fast local storage.
3Speed
If high-performance storage destinations are used for all layers, then all layers are quickly accessible, but the cost and space consumption of valuable local storage is increased
Solution Approach 1:
The patent applies local quality by providing high-performance storage only to frequently accessed layers while using lower-performance remote storage for less frequently accessed layers. This differential approach ensures that the limited local storage capacity is utilized efficiently, dedicating it to layers that require fast access while accepting slower access speeds for layers that are not frequently used.
Solution Approach 2:
The patent applies partial action by providing high-performance storage optimization only to the extent necessary for frequently accessed layers, rather than applying it to all layers. This partial optimization achieves the benefit of fast access speeds for critical operations while avoiding the excessive cost and space consumption that would result from applying high-performance storage to all layers uniformly.
Data Source
AI summary
Container-image layers can be managed. For example, a computing device can determine a first score for a first layer of a container image and a second score for a second layer of the container image. The computing device can determine that the first score corresponds to a first storage destination among several possible storage destinations. The computing device can also determine that the second score corresponds to a second storage destination among the possible storage destinations. The second storage destination can be different from the first storage destination. The computing device can then store (i) the first layer in the first storage destination based on the first layer being correlated to the first score, and (ii) the second layer in the second storage destination based on the second layer being correlated to the second score.


