Cloud Storage Provisioning Using ML Workload Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional storage systems face inefficiencies in data management and storage operations, particularly in handling flash drives without integrated storage controllers, leading to unnecessary write operations and reduced reliability.
Innovation Solution
A storage system architecture where the operating system initiates and controls data management processes, such as direct-mapped flash storage, to handle data blocks directly without address translation by storage controllers, offloading device management from flash drives and utilizing non-volatile RAM as a buffer for improved latency and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If storage controllers perform address translation and data management for flash drives, then data access efficiency is improved, but device complexity and redundant write operations increase
Solution Approach 1:
The patent extracts the storage controller from the flash drive, creating a controller-less flash drive architecture. The host system assumes the role of the storage controller, performing address translation and data management functions that were previously handled by dedicated hardware. This eliminates the need for complex onboard controllers while maintaining data access efficiency through software-based management.
Solution Approach 2:
The patent introduces a buffer memory as an intermediary between the flash drive and the host system. This buffer handles data temporarily during read/write operations, enabling the host to perform address translation and management functions without directly accessing the flash memory, thereby simplifying the flash drive architecture while maintaining efficient data access.
2Ease of operation
If flash drives include integrated storage controllers, then data management is simplified, but reliability decreases due to more write operations
Solution Approach 1:
By removing the storage controller from the flash drive, the patent eliminates the additional write operations that would otherwise be required to program the controller's internal memory and state machines. The host system performs all data management functions, reducing wear on the flash memory cells and improving reliability while maintaining ease of operation through software-based management.
Solution Approach 2:
The host system performs self-service by assuming the role of the storage controller. It directly manages data placement, address translation, and wear leveling algorithms, eliminating the need for a dedicated controller component. This approach simplifies the flash drive hardware while improving reliability by reducing the number of active components that could fail.
3Loss of information
If traditional storage controllers manage data blocks, then address translation is performed, but write operations increase and efficiency decreases
Solution Approach 1:
The patent extracts the address translation function from the storage controller and implements it in the host system's software. This allows for more flexible and efficient address management algorithms to be applied, improving storage operation efficiency while maintaining accurate address translation. The host can optimize translation operations based on workload patterns without being constrained by hardware limitations.
Solution Approach 2:
The patent implements preliminary action by having the host system pre-process data placement decisions and wear leveling operations before data is written to the flash memory. This proactive management approach optimizes write operations and extends flash memory lifespan without requiring a dedicated controller to manage these functions in real-time, thereby improving overall productivity.
Data Source
AI summary
Providing storage tailored for a storage consuming application, including: identifying, for an application that utilizes storage resources within a cloud-based storage system, one or more storage performance characteristics associated with the application; comparing the storage performance characteristics of the application that were identified with storage performance characteristics of storage resources of one or more cloud-based storage systems; and selecting, based on the comparing, one or more storage resources within the one or more cloud-based storage systems to provide storage services to the application.