Persistent Build Contexts for Parallel Reusable Pipeline Components
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional build pipelines in Continuous Integration/Continuous Delivery (CI/CD) processes are inefficient due to reliance on external inputs, create bridges between pipeline parts, and cannot parallelize tasks within a single pipeline instance, leading to increased delivery times and potential interference among steps.
Innovation Solution
The implementation of build contexts that encapsulate input and output data for each build step, allowing parallel execution and isolation of tasks within a single pipeline instance, enabling modular, scalable, and robust pipeline management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If build steps use statically configured inputs, then the pipeline structure is simple, but the pipeline cannot parallelize tasks and tasks must execute sequentially
Solution Approach 1:
The patent segments the build pipeline into independent build steps, each with its own isolated build context. This segmentation allows tasks to be executed in parallel because each step operates independently with its own context, eliminating the sequential dependency while maintaining manageable complexity through modular organization.
Solution Approach 2:
The patent introduces build contexts as intermediary structures that carry input and output data between build steps. These contexts act as mediators that enable parallel execution by providing isolated data containers for each step, allowing the pipeline to maintain structure while achieving parallelization through the intermediary context objects.
2Productivity
If build steps rely on external inputs, then the pipeline can share data between steps, but the pipeline becomes contingent on external source reliability and cannot parallelize
Solution Approach 1:
The patent extracts the input and output data from external sources and embeds them directly within isolated build contexts for each build step. This extraction eliminates dependency on external sources while maintaining data availability, enabling parallel execution without compromising reliability since each step has its own self-contained context.
Solution Approach 2:
The patent creates isolated copies of build contexts for each build step, allowing parallel execution of multiple steps with identical or different inputs. Each step receives a copy of the necessary data through its own context, enabling parallelization while ensuring reliability through data duplication rather than shared external dependencies.
3Productivity
If build steps create bridges between pipeline parts, then data can flow between steps, but parallel execution is prevented and delivery time increases
Solution Approach 1:
The patent segments data flow into isolated build contexts for each step, eliminating the need for bridges between pipeline parts. Each context is self-contained and can be processed independently in parallel, reducing delivery time while simplifying data flow management through clear boundaries between segments.
Solution Approach 2:
The patent transitions from a sequential, bridge-dependent data flow model to a parallel, context-based model by adding the dimension of isolation. Each build step operates in its own contextual dimension, allowing simultaneous execution without interference, thereby reducing delivery time while maintaining ease of operation through structured context management.
4Loss of time
If tasks must exit a long pipeline before other tasks can start, then sequential execution is simple, but delivery time drastically increases
Solution Approach 1:
The patent segments the pipeline into independent build steps with isolated contexts, allowing multiple segments to execute simultaneously rather than sequentially. This segmentation dramatically reduces delivery time by enabling parallel task execution while managing complexity through the modular context structure that organizes each segment independently.
Solution Approach 2:
The patent enables continuous parallel execution of build steps through isolated contexts, eliminating idle waiting time between sequential tasks. Each step can proceed continuously without waiting for previous steps to complete, maximizing resource utilization and reducing overall delivery time while maintaining organized pipeline orchestration through context management.
Data Source
AI summary
Techniques are provided for managing and isolating build process pipelines. An example method can include determining an identifier of a version commit event. The method can further include obtaining a build pipeline having a plurality of build steps comprising at least a first build step and a second build step based at least in part on the identifier of the version commit event. The method can further include executing the first build step of the build pipeline to access a first object from a repository. The method can further include generating a build context comprising at least a second object output based at least in part on the first build step. The method can further include executing the second build step of the build pipeline based at least in part on the build context, the build context accessible to the second build step.