Coordinator Daemon for Dynamic Container Cluster Management

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In a container-based distributed processing environment, newly generated containers lack existing cluster information, and when a distributed application fails, other nodes in the cluster cannot automatically detect and replace the down node, requiring manual administrator intervention.

Innovation Solution

A container-based distributed application management system that includes a coordinator daemon for storing and managing cluster setting information and agent daemons that monitor for issues, query the coordinator, and remove problematic applications, allowing new containers to join or replace failed ones dynamically.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a new container is dynamically generated to replace a failed distributed application, then system reliability is improved, but the new container lacks existing cluster information requiring manual administrator input

Engineering Contradiction:

Improvesystem reliabilityVSAvoidmanual administrator input

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The coordinator daemon stores cluster setting information in advance before any container failures occur. When a new container is generated, it automatically retrieves pre-stored cluster information from the coordinator daemon, eliminating the need for manual administrator input and enabling immediate integration into the cluster

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The coordinator daemon acts as an intermediary between the new container and the existing cluster. It stores and manages cluster setting information, automatically providing this information to new containers so they can join the cluster without manual configuration, thereby resolving the contradiction between reliability improvement and operational ease

Inventive Principle:

Principle #24Intermediary (Mediator)

2Difficulty of detecting and measuring

If agent daemons monitor distributed applications for failures, then detection capability is improved, but automatic sensing and replacement by other nodes is not achieved

Engineering Contradiction:

Improvedetection capabilityVSAvoidautomatic sensing and replacement

Core Design Contradiction:

Difficulty of detecting and measuringVSExtent of automation

Solution Approach 1:

Agent daemons continuously monitor distributed applications and provide feedback to the coordinator daemon about application status. When a failure is detected, the agent daemon notifies the coordinator, which then automatically triggers the generation of a new container with the appropriate cluster information, achieving automatic sensing and replacement

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system implements self-service through automatic failure detection and replacement. Agent daemons autonomously monitor applications, and when failures occur, the coordinator daemon automatically generates replacement containers with pre-stored cluster information, eliminating the need for manual intervention and achieving full automation

Inventive Principle:

Principle #25Self-service

3Manufacturing precision

If manual administrator input is required for cluster configuration, then configuration accuracy is improved, but system productivity and response time deteriorate

Engineering Contradiction:

Improveconfiguration accuracyVSAvoidsystem productivity

Core Design Contradiction:

Manufacturing precisionVSProductivity

Solution Approach 1:

Cluster setting information is stored in advance in the coordinator daemon with high configuration accuracy. When containers are dynamically generated, they automatically retrieve this pre-configured information, maintaining accuracy while eliminating manual input delays and significantly improving system productivity

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses copying by storing cluster setting information once in the coordinator daemon and automatically distributing copies to new containers. This ensures configuration accuracy is maintained while enabling rapid container deployment and improving overall system productivity

Inventive Principle:

Principle #26Copying

Data Source

PatentUS10572313B2Container-based distributed application management system and method

Publication Date: 2020.02.25 SAMSUNG SDS CO LTD

AI summary

A container-based distributed application management system includes a coordinator daemon configured to store and manage cluster setting information corresponding to a cluster formed by a plurality of distributed applications, the plurality of distributed applications configured to be separately executed in a plurality of containers; and a plurality of agent daemons configured to be separately installed and executed in the plurality of containers, and receive the cluster setting information from the coordinator daemon to execute at least one of the plurality of distributed applications installed in the plurality of containers, wherein the coordinator daemon and the plurality of agent daemons are implemented by using one or more hardware processors.