Decentralized Monitoring Agents for Zero-Downtime Software Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Centralized monitoring solutions for computer systems face high Total Cost of Ownership (TCO) due to large data storage needs and frequent code updates, which can cause downtime and inconvenience to end users.
Innovation Solution
Implementing a decentralized monitoring approach using monitoring agents that detect local system topology and store data locally, allowing for updates without code changes by shifting OLAP capabilities and data processing to the local agents, and utilizing a content-driven method with multiple content versions to manage updates seamlessly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a centralized monitoring solution stores monitoring data history for many different computer systems, then comprehensive monitoring coverage is improved, but data storage costs and Total Cost of Ownership increase
Solution Approach 1:
The patent segments the centralized monitoring architecture into distributed monitoring agents deployed at individual computer systems. Each agent locally collects and processes monitoring data, eliminating the need for a single centralized storage system to hold all data from all systems. This segmentation reduces total storage requirements while maintaining comprehensive monitoring coverage across the network.
2Adaptability or versatility
If monitoring code is updated frequently to support changes in database versions, operating systems, and hardware, then monitoring accuracy and compatibility are improved, but system availability deteriorates due to required re-initialization
Solution Approach 1:
The patent implements dynamic version management where multiple versions of monitoring agents can coexist and be switched between without system re-initialization. The system allows seamless transitions between agent versions, enabling frequent updates to maintain compatibility with changing infrastructure while preserving continuous system availability.
3Duration of action of stationary object
If monitoring data history is maintained for more than one year, then long-term behavior analysis capability is improved, but data storage costs and TCO increase
Solution Approach 1:
The patent applies local quality by allowing different retention periods for different types of monitoring data based on their analytical value. Frequently accessed recent data is retained locally at monitoring agents for immediate analysis, while older historical data can be archived or selectively retained. This differentiated approach enables long-term behavior analysis for critical data without uniformly retaining all data indefinitely, thereby reducing overall storage requirements.
4Adaptability or versatility
If the runtime system is re-initialized to utilize updated functionality, then software compatibility is improved, but end user convenience deteriorates due to temporary unavailability
Solution Approach 1:
The patent implements preliminary action by pre-loading and preparing updated monitoring agent versions in advance before they are needed. Multiple agent versions are maintained in readiness, allowing seamless switching to updated functionality without requiring system re-initialization or causing service interruptions. This preliminary preparation ensures software compatibility is maintained while preserving end user convenience.
Data Source
AI summary
A local monitoring system of a computer system to be monitored may receive a monitoring solution agent code portion and a first monitoring solution agent content portion. Version D may be assigned to the first content portion, and a status of version D may be set to active such that new end user sessions are initialized with a local agent comprising the code portion and version D. A second content portion may be uploaded and assigned to be version A. Responsive to an end user request, the status of version D may be set to ready and the status of version A may be set to active such that new sessions are initialized with an agent comprising the code portion and version A. A third content portion may then be uploaded and assigned to be version B. Responsive to an end user request, the status of version A may be to ready and the status of version B may be set to active such that new sessions are initialized with an agent comprising the code portion and version B. When sessions using version A no longer exist, version A may be deleted.