Code Commit Packaging for Selective CI/CD Rollback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing continuous integration/continuous deployment (CI/CD) systems lack a detailed process for identifying the reason behind a build or deployment failure and do not provide an automated fix, leading to prolonged downtime and difficulty in debugging, especially in low-code development platforms.
Innovation Solution
A code commit facility is introduced that generates discrete code change packages from failing code changes, deploys them in a target environment, detects failures, and rolls back specific files while maintaining others, thereby restructuring the deployment process to minimize downtime and enhance reporting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the entire code change is deployed as a single package, then deployment is simple, but if any file fails the entire deployment must be rolled back causing prolonged downtime
Solution Approach 1:
The patent segments the monolithic code change package into discrete code change packages, each containing one or more code change files. This allows selective deployment and rollback of individual packages rather than the entire code change, reducing downtime when failures occur while maintaining deployment simplicity through automated orchestration.
2Reliability
If discrete code change packages are deployed separately, then failed packages can be rolled back independently, but the deployment process becomes more complex
Solution Approach 1:
The code commit facility automatically performs the complex tasks of generating discrete code change packages, deploying them individually, detecting failures, and rolling back specific failed packages without human intervention. This automation maintains reliability through selective rollback while preventing deployment process complexity from burdening operators.
3Productivity
If traditional CI/CD systems detect a build failure, they stop the entire process without identifying the specific failing file, leading to prolonged analysis time
Solution Approach 1:
The patent segments the code change into discrete packages at the file level, allowing the system to identify which specific file or package caused the failure. This granular segmentation provides detailed failure diagnosis information while maintaining rapid deployment speed through automated continuous integration and deployment processes.
Solution Approach 2:
The code commit facility implements feedback mechanisms that detect failures at the discrete package level and provide specific information about which file or package failed. This feedback loop enables rapid identification of failure causes without stopping the entire deployment process, maintaining productivity while preventing information loss about failure diagnostics.
Data Source
AI summary
A code commit facility is provided operatively coupled to a continuous integration continuous deployment system to assist in addressing a failing code change during updating of a software application. The code commit facility includes generating discrete code change packages from the failing code change by separating independent code change files of the failing code change into the discrete code change packages. Further, the code commit facility deploys the discrete code change packages and based on the deploying, detects a failing discrete code change package. Based on detecting the failing discrete code change package, the code commit facility rolls back one or more code change files of the failing package, absent rolling back one or more other code change files of one or more other discrete code change packages to restructure the failing code change for deployment using the continuous integration continuous deployment system.