Concurrency Reduction via Publish-Subscribe Message Queue

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In microservices architecture, multiple instances of the same service concurrently requesting data from a back-end service lead to redundant requests, causing unnecessary stress and delays, as existing technologies fail to effectively prevent redundant traffic.

Innovation Solution

Implementing a concurrency reduction message service using publish-subscribe patterns, where services subscribe to a shared message queue, map requests to keys, cache responses, and block redundant requests until a response is received, thereby reducing concurrency and preventing redundant requests to the back-end service.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If multiple service instances concurrently request data from a back-end service, then system productivity is improved through parallel processing, but redundant requests cause increased stress and delays on the back-end service

Engineering Contradiction:

Improvesystem productivityVSAvoidback-end service reliability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system performs preliminary actions by checking for existing requests before allowing new concurrent requests to proceed. Each service instance checks the request queue and cache before submitting a new request, preventing redundant requests from being sent to the back-end service while maintaining the ability to process multiple requests in parallel when necessary

Inventive Principle:

Principle #10Preliminary action

2Productivity

If service instances redundantly request the same data multiple times, then system productivity is maintained through continuous processing, but memory consumption and processing time increase

Engineering Contradiction:

Improvesystem productivityVSAvoidprocessing time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system checks the request queue and cache before allowing new requests to proceed. This preliminary check prevents redundant requests from being processed, eliminating unnecessary processing time and memory consumption while maintaining system productivity through efficient request handling

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms where service instances are notified when requests are completed or when data is available in the cache. This feedback loop allows service instances to stop waiting and proceed with alternative operations, reducing overall processing time for redundant requests

Inventive Principle:

Principle #23Feedback

3Reliability

If a proxy or cache service is introduced to reduce concurrency, then back-end service stress is reduced, but device complexity and implementation difficulty increase

Engineering Contradiction:

Improveback-end service reliabilityVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system enables service instances to self-manage concurrency by implementing the check-then-proceed logic within each service instance itself. Each service instance independently checks the request queue and cache before submitting requests, eliminating the need for a centralized proxy or cache management service and reducing overall system complexity

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS11323534B2Concurrency reduction through publish-subscribe patterns

Publication Date: 2022.05.03 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Receiving a plurality of concurrent incoming requests from the plurality of services, with each incoming request and each service subscribing to a topic in a shared message queue. The queue is used for: notifying other subscribers when a request to a back-end service is being made; and/or notifying other subscribers when a response from the back-end service is received.