Global Node Architecture for Read-Heavy Applications

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Global application performance deteriorates with distance from the primary service, leading to slower performance in regions like Asia Pacific and Europe, Middle East, and Africa, and existing solutions like WAN accelerators are not sufficiently effective, resulting in high latency and operational costs due to divergent regional instances and database replication.

Innovation Solution

A global node system architecture with self-contained nodes that replicate data and synchronize updates, automatically detect the nearest node for read operations, and route requests to minimize latency, eliminating the need for additional infrastructure like WAN accelerators and reducing operational costs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If data is stored at a primary service location, then data consistency is maintained, but access latency increases for distant global regions

Engineering Contradiction:

Improvedata consistencyVSAvoidaccess latency

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system segments the centralized data storage into multiple distributed global nodes, each holding a replica of the data. This allows users in different geographical regions to access data from the nearest node, reducing latency while maintaining consistency through synchronization mechanisms.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent creates exact replicas of the data across multiple global nodes distributed throughout the world. Each node stores a complete copy of the data, enabling local access without requiring remote connections to the primary service, thereby reducing access latency while preserving data consistency.

Inventive Principle:

Principle #26Copying

2Productivity

If WAN accelerators are deployed to improve global performance, then network throughput increases, but system complexity and operational costs increase

Engineering Contradiction:

Improvenetwork throughputVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent extracts the need for WAN accelerators by implementing data replication at the source (global nodes). Instead of adding network acceleration infrastructure, the system directly addresses performance by placing data copies closer to users, eliminating the need for complex acceleration devices.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The global nodes act as intermediaries between users and the primary service. Each node serves as a local cache that fulfills read requests without requiring communication with the primary service or other nodes, simplifying the system architecture compared to WAN accelerator approaches.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Speed

If independent regional instances are deployed, then local performance improves, but data divergence and functional inconsistency occur

Engineering Contradiction:

Improvelocal performanceVSAvoiddata consistency

Core Design Contradiction:

SpeedVSReliability

Solution Approach 1:

The patent merges the benefits of distributed performance with centralized consistency by having all global nodes maintain identical data replicas. Unlike independent regional instances that diverge, these nodes are synchronized to hold the same data, combining local speed with global consistency.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

All global nodes maintain homogeneous data states by storing exact replicas of the same dataset. This homogeneity ensures that any node can serve any request with consistent data, preventing the functional divergence seen in independent regional instances while maintaining local performance.

Inventive Principle:

Principle #33Homogeneity

4Reliability

If database replication is implemented for global reach, then data availability improves, but operational and developmental costs increase significantly

Engineering Contradiction:

Improvedata availabilityVSAvoidoperational costs

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent uses standard, commodity server infrastructure for global nodes rather than expensive specialized database replication infrastructure. Each node is a simple, self-contained server that can be deployed using off-the-shelf hardware and software, significantly reducing operational and developmental costs while achieving global data availability.

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

Solution Approach 2:

Each global node operates as a self-contained, autonomous unit that independently serves local requests without requiring complex coordination or management infrastructure. This self-service capability eliminates the need for expensive centralized management systems and reduces operational overhead.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS10506038B1Method and system for implementing a global node architecture

Publication Date: 2019.12.10 JPMORGAN CHASE BANK NA

AI summary

According to an embodiment of the present invention, a global node system architecture that is specific to a read many and write infrequently application comprising: a plurality of global nodes distributed throughout a global region, each global node comprises: a memory that stores a set of data that is replicated at each node of the plurality of global nodes where each node is a self-contained node so that each node of the plurality of global nodes comprises an exact replica of data; and a processor coupled to the memory and programmed to: synchronize data updates at each of the plurality of global nodes; automatically detect a nearest global node with respect to each of the plurality of global nodes; and initiating a communication with the nearest global node.