Layered Edge Storage With Erasure Coding for Atomic Access

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Edge computing systems face challenges in reducing the cost of operation of backend layer servers, maintaining consistency in data access, and minimizing latency, particularly in scenarios where edge servers lack sufficient storage capacity and resources.

Innovation Solution

A layered distributed storage (LDS) system is introduced, which employs a two-layer erasure-coded fault-tolerant distributed storage architecture. This system enables atomic access for read and write operations, optimizes communication and storage costs, and ensures fault-tolerance by using erasure codes such as minimum bandwidth regenerating (MBR) codes and random linear network codes (RLNC).

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If edge servers store all files locally to provide fast access, then read latency is reduced, but storage capacity requirements increase significantly

Engineering Contradiction:
Improveread latencyVSAvoidstorage capacity
Core Design Contradiction:
SpeedVSQuantity of substance

Solution Approach 1:

The system divides the storage architecture into two segments: edge layer servers for frequently accessed files and backend layer servers for permanent storage. This segmentation allows read operations to be served from the fast edge layer when possible, while maintaining comprehensive storage capacity in the backend layer, thus resolving the contradiction between read latency and storage capacity requirements.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a vertical dimension to the storage hierarchy by adding the backend layer above the edge layer. This dimensional expansion creates a multi-layer storage architecture where data can be distributed across different levels based on access frequency, enabling fast local access for hot data while maintaining total storage capacity in the backend, thereby resolving the storage capacity vs. latency tradeoff.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Loss of energy

If edge servers have limited storage capacity, then operational cost is reduced, but ability to store frequently accessed files locally deteriorates

Engineering Contradiction:
Improveoperational costVSAvoiddata access speed
Core Design Contradiction:
Loss of energyVSSpeed

Solution Approach 1:

The backend layer servers act as an intermediary storage layer between edge servers and permanent storage. When edge servers lack capacity for frequently accessed files, the system can retrieve these files from the backend layer, providing a middle ground that balances operational cost constraints with the need for fast data access, thus resolving the contradiction between operational cost and data access speed.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If strong consistency mechanisms are implemented across distributed servers, then data consistency is improved, but communication overhead and latency increase

Engineering Contradiction:
Improvedata consistencyVSAvoidcommunication latency
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system implements local quality by providing strong consistency guarantees specifically for the edge layer where it is most needed for client operations, while using erasure coding for backend storage which provides adequate consistency with lower communication overhead. This differentiated approach to consistency based on location in the hierarchy resolves the contradiction between data consistency and communication latency.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent changes the consistency parameter from strong consistency to erasure coding for backend storage operations. Erasure coding provides sufficient data integrity and recovery capabilities with significantly reduced communication overhead compared to traditional strong consistency protocols, thus resolving the contradiction between data consistency reliability and communication latency.

Inventive Principle:
Principle #35Parameter changes

4Quantity of substance

If backend layer servers are used for permanent storage, then storage capacity is increased, but communication cost between layers increases

Engineering Contradiction:
Improvestorage capacityVSAvoidcommunication cost
Core Design Contradiction:
Quantity of substanceVSLoss of energy

Solution Approach 1:

The system performs preliminary actions by pre-coding data using erasure coding before transferring it to the backend layer. This preliminary encoding reduces the amount of data that needs to be communicated and stored redundantly, thereby decreasing communication cost between layers while maintaining adequate storage capacity in the backend, thus resolving the contradiction between storage capacity and communication cost.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3631641B1Layered distributed storage system and techniques for edge computing systems
Publication Date: 2025.01.22 MASSACHUSETTS INST OF TECH
  • EP3631641B1 patent drawingFigure 1
  • EP3631641B1 patent drawingFigure 1A
  • EP3631641B1 patent drawingFigure 2~2A

AI summary

A two-layer erasure-coded fault-tolerant distributed storage system offering atomic access for read and write operations is described. In some embodiments, a class of erasure codes known as regenerating codes (e.g. minimum bandwidth regenerating codes) for storage of data in a backend layer is used to reduce the cost of backend bulk storage and helps reduce communication cost of read operations, when a value needs to be recreated from persistent storage in the backend layer. By separating the functionality of edge layer servers and backend servers, a modular implementation for atomicity using storage-efficient erasure-codes is provided. Such a two-layer modular architecture permits protocols needed for consistency implementation to be substantially limited to interaction between clients and the edge layer, while protocols needed to implement erasure codes are substantially limited to the interaction between edge and backend layers.