Node Warming in Distributed Data Store Replication

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Distributed data stores face interruptions and reduced data redundancy when a node becomes unavailable, particularly in master-slave replication, and reduced copy numbers in master-master replication during node replacement.

Innovation Solution

A method that identifies a source node storing client data, configures a new node as a slave to receive streamed data from the source node, and then transitions the new node to a master node, maintaining data redundancy and service continuity.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If master-slave replication is used to ensure data redundancy, then data reliability is improved, but service interruption occurs during node failure election

Engineering Contradiction:
Improvedata reliabilityVSAvoidservice continuity
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent applies preliminary action by pre-configuring a new node with data from an available source node before the failed node is fully replaced. The new node is configured as a slave to a living node and pre-loaded with data through streaming, so that when it becomes a master, service continuity is maintained without interruption.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If master-master replication is used to maintain service continuity, then service productivity is improved, but data redundancy is reduced when a new node is added

Engineering Contradiction:
Improveservice continuityVSAvoiddata redundancy
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent resolves this contradiction by performing preliminary data streaming to the new node before it is fully activated as a master. The new node is configured as a slave to an available node and pre-loaded with data, ensuring that when it becomes a master, full data redundancy is restored without service interruption.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent uses an intermediary approach by introducing a transitional slave state for the new node. During this intermediate phase, the new node receives data from a source node as a slave, acting as a mediator that gradually builds up data redundancy before transitioning to full master status, thus maintaining both service continuity and data redundancy.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Productivity

If a new node is added to replace a failed node in master-master replication, then service continuity is maintained, but the number of data copies is reduced

Engineering Contradiction:
Improveservice continuityVSAvoidnumber of data copies
Core Design Contradiction:
ProductivityVSQuantity of substance

Solution Approach 1:

The patent applies preliminary action by pre-populating the new node's data store with data from a source node before the new node fully joins the cluster as a master. This is achieved by configuring the new node as a slave to an available node and streaming data in advance, ensuring data copies are restored before the node becomes operational as a master.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent uses copying by streaming data from a source node's data store to the new node's data store. The new node receives and stores copies of the client data that was previously stored by the failed node, thereby restoring the number of data copies to the original level.

Inventive Principle:
Principle #26Copying

Data Source

PatentUS10749921B2Techniques for warming up a node in a distributed data store
Publication Date: 2020.08.18 NETFLIX INC
  • US10749921B2 patent drawing
  • US10749921B2 patent drawing
  • US10749921B2 patent drawing

AI summary

In various embodiments, a node manager configures a “new” node as a replacement for an “unavailable” node that was previously included in a distributed data store. First, the node manager identifies a source node that stores client data that was also stored in the unavailable node. Subsequently, the node manager configures the new node to operate as a slave of the source node and streams the client data from the source node to the new node. Finally, the node manager configures the new node to operate as one of multiple masters nodes in the distributed data store. Advantageously, by configuring the node to implement a hybrid of a master-slave replication scheme and a master-master replication scheme, the node manager enables the distributed data store to process client requests without interruption while automatically restoring the previous level of redundancy provided by the distributed data store.