Storage System Throughput-Based Synchronous Replication Recovery

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional data replication methods face challenges in resuming synchronous replication mode after suspension, leading to potential network congestion issues and performance fluctuations, which can impact high availability and consistency in storage systems.

Innovation Solution

Implementing throughput-based timing for synchronous replication recovery, where storage systems maintain and update throughput statistics during asynchronous replication cycles to optimize the transition back to synchronous mode, avoiding aggressive or slow recovery that could disrupt host device performance and data consistency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If synchronous replication is resumed immediately after suspension, then data consistency between source and target storage systems is improved, but network congestion and host device performance fluctuations occur

Engineering Contradiction:

Improvedata consistencyVSAvoidhost device performance

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system dynamically adjusts the replication mode transition timing based on real-time network throughput conditions. Instead of using a fixed or aggressive resume schedule, the system monitors network performance metrics and adapts the resumption timing to current conditions, allowing optimal balance between data consistency and host performance across varying operational states

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system implements feedback mechanisms by monitoring network throughput statistics and using this information to control the transition from asynchronous to synchronous replication mode. The throughput statistics serve as feedback that informs whether conditions are favorable for resuming synchronous replication, preventing premature resumption that would cause performance degradation

Inventive Principle:

Principle #23Feedback

2Productivity

If synchronous replication recovery is delayed, then host device performance is preserved, but source and target synchronization loss increases adversely impacting RPO and HA goals

Engineering Contradiction:

Improvehost device performanceVSAvoidsource and target synchronization

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system dynamically determines the optimal moment to resume synchronous replication by continuously monitoring network throughput. This dynamic approach ensures that recovery actions are taken at the most favorable moment - not too early to avoid performance issues, and not too late to prevent excessive synchronization loss, thereby maintaining both host performance and data consistency

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the operational parameter of replication mode based on throughput statistics. When throughput metrics indicate favorable network conditions, the system transitions from asynchronous to synchronous mode. This parameter change allows the system to optimize the balance between performance preservation and synchronization maintenance based on actual network conditions

Inventive Principle:

Principle #35Parameter changes

3Reliability

If aggressive synchronous replication recovery is implemented, then RPO and HA goals are met, but network congestion issues and configuration errors such as bouncing links are exacerbated

Engineering Contradiction:

ImproveRPO and HA goalsVSAvoidnetwork congestion and configuration errors

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The system takes preliminary action by monitoring network throughput statistics before attempting to resume synchronous replication. This advance monitoring allows the system to detect unfavorable network conditions and prevent recovery actions that would exacerbate network congestion or trigger configuration errors, thereby avoiding the harmful effects before they occur

Inventive Principle:

Principle #9Preliminary anti-action

Solution Approach 2:

The system provides cushioning against network instability by using throughput statistics to buffer the recovery process. Instead of immediately resuming synchronous replication after suspension, the system waits for throughput metrics to indicate stable network conditions, effectively cushioning the system against the potential harmful effects of premature recovery attempts

Inventive Principle:

Principle #11Beforehand cushioning (Prior cushioning)

Data Source

PatentUS11099767B2Storage system with throughput-based timing of synchronous replication recovery

Publication Date: 2021.08.24 EMC IP HLDG CO LLC

AI summary

An apparatus includes at least one processing device comprising a processor coupled to a memory, with the processing device being configured to maintain throughput statistics for at least portions of one or more asynchronous replication cycles of a replication process carried out between a first storage system and a second storage system, and to control a transition of the replication process from an asynchronous replication mode to a synchronous replication mode based at least in part on the maintained throughput statistics. The throughput statistics are illustratively updated in conjunction with each of a plurality of detected suspensions of the synchronous replication mode, and the asynchronous replication mode is reinitiated responsive to each of the detected suspensions of the synchronous replication mode. Other embodiments include methods and computer program products.