Load Balanced Computer Maintenance Automation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for maintaining load balanced computers often require service unavailability, are difficult to schedule due to unpredictable workloads, and involve risks such as request failures and increased maintenance times, especially when handling a large group of computers.
Innovation Solution
The implementation of a system that automates maintenance by identifying computers requiring maintenance, updating them without removing them from the load balancing group, and ensuring minimal disruption by using a maintenance agent to manage updates and reintegration, thus avoiding the need for additional resources and minimizing service downtime.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a computer is removed from the load balancing scheme for maintenance, then maintenance can be performed on the computer, but the service becomes unavailable and pending requests must be terminated
Solution Approach 1:
The system dynamically adjusts the maintenance process by monitoring request termination status and automatically controlling when to remove and re-add computers to the load balancing scheme. The maintenance window is not fixed but adapts based on the actual state of pending requests, allowing the system to maintain service availability while performing necessary maintenance.
Solution Approach 2:
The load balancer provides feedback about the status of pending requests to the maintenance management system. This feedback mechanism allows the system to determine when all pending requests have been terminated and when it is safe to re-add the computer to the load balancing scheme, ensuring service availability is restored as quickly as possible.
2Ease of manufacture
If a maintenance window is created to maintain load balanced computers, then maintenance can be performed on all computers simultaneously, but the service is unavailable and requests may fail
Solution Approach 1:
Instead of maintaining all computers simultaneously during a single maintenance window, the system segments the maintenance process by handling computers individually or in small groups. The load balancer manages the removal and re-addition of computers one at a time, ensuring that maintenance can be performed on multiple computers without causing complete service unavailability.
Solution Approach 2:
The system performs preliminary actions by automatically terminating pending requests before removing computers from the load balancing scheme. This preliminary termination of requests ensures that no requests are lost or failed during the maintenance process, and the computer can be safely taken offline for maintenance.
3Ease of manufacture
If computers are manually removed from the load balanced group for maintenance, then maintenance can be performed, but there is risk of unanticipated longer maintenance times and service disruption
Solution Approach 1:
The maintenance management system automatically monitors the duration of maintenance operations and can trigger alerts or actions if maintenance takes longer than expected. The system self-manages the entire process from removing computers to monitoring maintenance duration and re-adding them, reducing the need for manual intervention and minimizing service disruption time.
4Ease of manufacture
If the load balancer stops issuing processing requests to a computer for maintenance, then the computer can be maintained, but the time to terminate processes is significant
Solution Approach 1:
The system performs preliminary termination of pending requests using the feedback mechanism before physically removing the computer from the load balancing scheme. This preliminary action ensures that the computer is already in a safe state for maintenance, reducing the actual time the computer needs to be offline and minimizing process termination time.
Data Source
AI summary
The present invention is directed towards methods and systems for performing maintenance on computers included in a group of load balanced computer. A maintenance agent identifies a computers in the group as requiring maintenance, and a second computer as having up-to-date maintenance and a load level below a limit. The load balancer may assign a request to the second computer responsive to the identification. Responsive to a determination that spare capacity from the plurality of computers available to receive requests is above the threshold and that the first computer is idle, the maintenance agent may remove, for maintenance, the first computer from the plurality of computers available to receive requests. The maintenance agent may identify the first computer as having up-to-date maintenance responsive to the maintenance.