I/O Score-Based Hybrid Replication in Storage Systems

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional data replication systems face challenges in efficiently managing replication processes, leading to increased journal lag, resource consumption, and prolonged recovery times due to the need for validating point-in-time data replicas, which can result in reduced performance and longer recovery objectives.

Innovation Solution

A method and system that determine properties of each input/output (I/O) operation and adapt replication thresholds based on operating conditions, allowing for dynamic selection between continuous replication and snapshot replication, optimizing replication strategies in real-time based on I/O properties and system conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If continuous replication is performed for all I/O operations, then data integrity is improved, but resource consumption increases

Engineering Contradiction:
Improvedata integrityVSAvoidresource consumption
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The system dynamically changes the replication parameter (continuous vs. snapshot) based on I/O score thresholds and operating conditions, transitioning between replication modes to balance data integrity and resource consumption

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The replication strategy is made dynamic by continuously monitoring operating conditions and I/O properties, allowing the system to adapt between continuous and snapshot replication based on real-time scores and thresholds

Inventive Principle:
Principle #15Dynamics

2Reliability

If validation of point-in-time data replicas is performed, then reliability is improved, but recovery time increases

Engineering Contradiction:
Improvedata validationVSAvoidrecovery time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

Instead of validating all replicas, the system performs partial validation by selecting only critical I/O operations for continuous replication based on scoring, reducing validation time while maintaining reliability for important data

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The system performs preliminary assessment of I/O operations by scoring them before replication, identifying which operations require validation and which can use snapshot replication, thereby reducing overall recovery time

Inventive Principle:
Principle #10Preliminary action

3Use of energy by moving object

If snapshot replication is used instead of continuous replication, then resource consumption is reduced, but journal lag increases

Engineering Contradiction:
Improveresource consumptionVSAvoidjournal lag
Core Design Contradiction:
Use of energy by moving objectVSLoss of time

Solution Approach 1:

The system changes the replication parameter dynamically based on I/O scores and operating conditions, switching between snapshot and continuous replication modes to balance resource consumption and journal lag

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The replication approach transitions from static to dynamic by continuously evaluating I/O properties and system conditions, adapting the replication strategy to minimize both resource usage and journal lag in real-time

Inventive Principle:
Principle #15Dynamics

4Reliability

If replication threshold is lowered to increase continuous replication, then data integrity is improved, but productivity decreases

Engineering Contradiction:
Improvedata integrityVSAvoidreplication efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The system optimizes the replication threshold parameter dynamically based on operating conditions, adjusting it to achieve the best balance between data integrity and replication efficiency rather than using a fixed low threshold

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The replication threshold becomes a dynamic parameter that adapts to system conditions and I/O properties, allowing the system to maintain high data integrity when necessary while preserving productivity during normal operations

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS10467102B1I/O score-based hybrid replication in a storage system
Publication Date: 2019.11.05 EMC IP HLDG CO LLC
  • US10467102B1 patent drawing
  • US10467102B1 patent drawing
  • US10467102B1 patent drawing

AI summary

Described embodiments provide systems and methods for operating a storage system that may determine one or more properties for each of a plurality of input/output (I/O) to a production volume of a storage system and monitor one or more operating conditions of the storage system. Embodiments may further include determining a score for each I/O based upon one or more of: the one or more properties of the I/O and the one or more operating conditions; adapting a replication threshold based upon the one or more operating conditions; comparing the determined score for each I/O to the adapted replication threshold; and based upon the comparison, performing continuous replication or snapshot replication for each I/O.