Multicast ACK Management in Distributed Cache Servers

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In distributed cache environments, multicast reliability is challenging due to the need for managing acknowledgments from a large number of receivers, leading to excessive network load, CPU and memory resource consumption, and long response times if inefficient algorithms are used.

Innovation Solution

A multicasting method where a distributed cache server transmits messages through multicast, registers UNACK messages in an unacknowledged message map, and stops additional transmissions when a predetermined number of UNACK messages is reached, while also managing ACK messages by identifying the largest message number for each client node and automatically confirming smaller ACK messages as received, thereby reducing the number of ACK messages transmitted.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If multicast is used to transmit messages to a large number of client nodes, then communication efficiency is improved, but reliability becomes difficult to guarantee due to the complexity of managing acknowledgments from all receivers

Engineering Contradiction:

Improvecommunication efficiencyVSAvoidmulticast reliability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent segments the acknowledgment management into two parts: (1) The server segments the list of client nodes into multiple groups, and (2) ACK messages are segmented and sent to different gateway nodes based on the UNACK message map. This segmentation reduces the complexity of managing acknowledgments from all receivers simultaneously while maintaining reliable delivery tracking.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces gateway nodes as intermediaries between the multicast server and client nodes. The gateway nodes receive and forward ACK messages to the server, acting as mediators that simplify the server's acknowledgment management. The server only needs to communicate with gateway nodes rather than managing individual client acknowledgments directly.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If acknowledgments from all receivers are managed to guarantee reliability, then multicast reliability is improved, but excessive network load and CPU/memory resource consumption occur

Engineering Contradiction:

Improvemulticast reliabilityVSAvoidnetwork load and computing resource consumption

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The patent extracts the acknowledgment management function from the server and relocates it to dedicated gateway nodes. The server extracts only the essential tracking function (maintaining UNACK message map) while gateway nodes handle the actual ACK message forwarding. This extraction reduces the server's processing burden and network load.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent implements partial acknowledgment management by tracking only unacknowledged messages in the UNACK message map rather than managing all acknowledgments explicitly. The server sends multicast messages and partially tracks only those that need acknowledgment, reducing the overall acknowledgment management overhead while maintaining necessary reliability.

Inventive Principle:

Principle #16Partial or excessive action

3Reliability

If acknowledgments from all receivers are managed to guarantee reliability, then multicast reliability is improved, but long response time occurs due to the complexity of acknowledgment processes

Engineering Contradiction:

Improvemulticast reliabilityVSAvoidresponse time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent performs preliminary action by pre-establishing the UNACK message map that tracks which client nodes need to acknowledge which messages. This preliminary tracking structure is set up before acknowledgment processing begins, allowing for efficient lookup and forwarding of ACK messages through gateway nodes without complex real-time decision-making.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

Gateway nodes act as intermediaries that handle ACK message forwarding, reducing the response time burden on the server. The server quickly records UNACK messages and relies on gateway nodes to efficiently manage the actual ACK message transmission and tracking, reducing overall system response time while maintaining reliability.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS10701170B2Multicasting method in distributed cache environment, and distributed cache server using the same

Publication Date: 2020.06.30 TMAXSOFT

AI summary

A multicasting method in a distributed cache environment is provided. The multicasting method includes steps of: (a) a specific distributed cache server among a plurality of distributed cache servers in the distributed cache environment, transmitting certain messages to client nodes including at least one another distributed cache server among the plurality of distributed cache servers in the distributed cache environment through multicast and then sequentially registering, in an unacknowledged message map, UNACK messages representing messages without their corresponding ACK messages from their corresponding client nodes among the certain messages; and (b) the specific distributed cache server checking the number of the UNACK messages registered in the unacknowledged message map and, if the number reaches a first predetermined number, stopping transmission of additional messages to the client nodes through the multicast.