Backup Storage Node Replication via Local Change Data Caching

Solution Overview

Problem

Conventional storage platforms experience performance slowdowns due to repetitive and time-consuming data transmissions between storage nodes in replication relationships, particularly when updating mirrored volumes.

Innovation Solution

By initially receiving service requests for changing mirrored volumes at the backup storage node, which retains a copy of the change data, the primary storage node can apply changes locally and send only replication instructions to the backup node, eliminating the need for retransmitting full change data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the primary storage node sends all change data to the backup storage node for every primary volume change, then the backup volume maintains accurate replication, but network bandwidth is consumed and transmission time increases

Engineering Contradiction:

Improvereplication accuracyVSAvoidnetwork bandwidth consumption

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The patent extracts only the essential replication information (metadata, change identifiers, location information) from the full change data, separating the replication control signals from the actual data payload. This allows the backup node to perform replication without receiving complete data copies, significantly reducing network bandwidth consumption while maintaining replication accuracy.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The backup storage node creates and maintains a local copy of the change data in its buffer memory, eliminating the need to receive repeated copies of the same data from the primary node. This local copying mechanism reduces network traffic while ensuring the backup volume can be updated accurately.

Inventive Principle:

Principle #26Copying

2Reliability

If the primary storage node sends all change data to the backup storage node for every primary volume change, then the backup volume maintains accurate replication, but transmission time increases and performance slows down

Engineering Contradiction:

Improvereplication accuracyVSAvoiddata transmission time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent extracts only the essential replication information (metadata, change identifiers, location information) from the full change data, separating the replication control signals from the actual data payload. This allows the backup node to perform replication without receiving complete data copies, significantly reducing network bandwidth consumption while maintaining replication accuracy.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The backup storage node proactively maintains a buffer copy of change data in its local memory, performing the data retrieval action in advance before it is needed for replication. This preliminary action eliminates waiting time during the replication process, as the backup node can immediately apply changes using its local copy rather than requesting data from the primary node.

Inventive Principle:

Principle #10Preliminary action

3Reliability

If service requests are received at the primary storage node and then forwarded to the backup node with full change data, then replication is maintained, but unnecessary data retransmission occurs

Engineering Contradiction:

Improvereplication maintenanceVSAvoidservice request completion speed

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The backup storage node creates and maintains a local copy of the change data in its buffer memory, eliminating the need to receive repeated copies of the same data from the primary node. This local copying mechanism reduces network traffic and accelerates service request completion, as the backup node can immediately process replication requests using its cached data.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The backup storage node proactively maintains a buffer copy of change data in its local memory, performing the data retrieval action in advance before it is needed for replication. This preliminary action eliminates waiting time during the replication process, as the backup node can immediately apply changes using its local copy rather than requesting data from the primary node.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20250138959A1Efficient data transmissions between storage nodes in replication relationships

Publication Date: 2025.05.01 NVIDIA CORP

AI summary

A storage platform (100) improves data flow when modifying mirrored volumes. A backup storage component (120A) that receives a service request keeps a copy of change data when redirecting the service request to a primary storage component (120B) that owns the volume that the service request targets. The primary storage (120B) component does not need to return the change data to the backup storage component (120A) when the primary storage component (120B) instructs the backup storage component (120A) to apply the modification request to the backup copy of the volume.