Configuration Change Control via Exposure State Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Organizations managing large computing assets or providing configuration items face a trade-off between roll-out speed and risk management, as updated configuration items may contain bugs or flaws, potentially leading to wide-scale outages.
Innovation Solution
A solution that involves receiving a second configuration item for displacement in an exposure group, determining an exposure state to split the group into portions continuing with the first configuration item and receiving the second configuration item, and deploying the second configuration item in accordance with the exposure state, with updates based on trigger events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If updated configuration items are deployed to all users immediately, then roll-out speed is improved, but the risk of wide-scale outages increases due to potential bugs or flaws
Solution Approach 1:
The user base is segmented into multiple exposure groups that receive configuration item updates at different rates. The system divides the population into cohorts, with each cohort receiving the update at a staged pace rather than all at once, allowing risk containment while maintaining deployment momentum
Solution Approach 2:
The exposure state is dynamically adjusted based on system performance monitoring and trigger events. The configuration item exposure percentage changes over time according to observed stability, allowing the system to accelerate deployment when stable or slow down/rollback when issues are detected
2Reliability
If configuration changes are rolled out gradually to limit risk exposure, then system stability is improved, but roll-out speed decreases
Solution Approach 1:
The system continuously monitors performance metrics and system stability during the rollout process. Trigger events based on performance thresholds provide feedback that automatically adjusts the exposure state, enabling the system to learn from real-world performance and optimize the rollout pace accordingly
Solution Approach 2:
The exposure percentage parameter is changed dynamically during deployment based on observed system stability and performance metrics. The system adjusts this parameter upward when stability is maintained and downward when issues arise, optimizing the balance between speed and risk
Data Source
AI summary
Solutions for balancing speed and risk by managing configuration changes include: receiving a second configuration item for displacement, in an exposure group, of a first configuration item; receiving an exposure state, wherein the exposure state indicates an exposure tree comprising a first configuration item branch and a second configuration item branch; determining, based at least on the exposure state: a first portion of the exposure group to continue with the first configuration item, and a second portion of the exposure group to receive the second configuration item; deploying the second configuration item to the second portion of the exposure group, in accordance with the exposure state; receiving, from the central orchestrator, an updated exposure state; and deploying the second configuration item in accordance with the updated exposure state. In some examples, the exposure tree is a hierarchical binary tree. An exemplary configuration item includes a software application version.