Direct-Mapped Flash Storage via Host-Side Address Translation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems face inefficiencies in data management and storage operations, particularly in handling flash storage devices without integrated storage controllers, leading to redundant processes and reliability issues.
Innovation Solution
The implementation of a direct-mapped flash storage system that addresses data blocks directly without address translation by the storage controllers, allowing the operating system to initiate and control storage processes, and utilizing non-volatile solid state storage units with embedded controllers for efficient data management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If storage controllers perform address translation for flash storage devices, then data access flexibility is improved, but system complexity and redundant operations increase
Solution Approach 1:
The patent extracts the address translation function from the storage controller and implements it directly in the host processor. This removes the redundant address translation layer in the storage controller, simplifying the storage system architecture while maintaining data access flexibility through host-side management.
Solution Approach 2:
The host processor serves itself by performing address translation operations that were previously offloaded to the storage controller. This self-service approach eliminates the need for complex controller-based translation mechanisms while maintaining the required adaptability in data access patterns.
2Ease of operation
If storage systems use integrated storage controllers, then data management control is improved, but operational reliability deteriorates due to single point of failure
Solution Approach 1:
The patent segments the storage management functions by separating the control plane (host processor) from the data plane (storage devices). This segmentation removes the single point of failure represented by the integrated storage controller, as the host processor continues to manage data even if individual storage devices fail.
Solution Approach 2:
The host processor acts as an intermediary between the host system and storage devices, maintaining data management control without requiring a dedicated storage controller. This intermediary role improves reliability by distributing control functions and eliminating the vulnerable integrated controller architecture.
3Reliability
If storage systems perform proactive data rebuilding, then data reliability is improved, but storage operation time increases
Solution Approach 1:
The patent implements preliminary action by performing data rebuilding proactively before failures occur or during low-utilization periods. The host processor initiates rebuild operations in advance, populating spare capacity and reconstructing data across multiple devices before actual failures impact system operation, thus reducing the time penalty of reactive rebuilding.
Data Source
AI summary
An illustrative method includes receiving, by a container storage interface (CSI) filter driver executing on a node of a cluster managed by a container orchestrator, a request to deploy a containerized application; selecting, by the CSI filter driver based on one or more attributes of the containerized application, a storage system from a plurality of storage systems attached to a plurality of clusters managed by the container orchestrator; and transmitting, by the CSI filter driver, a command to a CSI driver executing on a node of a cluster to which the storage system is attached, the cluster included in the plurality of clusters, the command configured to direct the CSI driver to provision a volume on the storage system for use with the containerized application.