SE-Radio Episode 338: Brent Laster on the Jenkins 2 Build Server

Syntax Choices

The debate between declarative and imperative syntax in Jenkins pipelines is pivotal for developers. Brent Laster explains that scripted syntax offers flexibility, allowing programmers to use constructs like loops and variables, which is ideal for those with a programming background 1. In contrast, declarative syntax provides a structured approach, making it more accessible for users familiar with traditional Jenkins setups 1. This structured nature aids in error checking and integrates well with Jenkins' GUI, Blue Ocean 1.

Often in the case with declarative programming, 5% of the time you just have to write a little bit of imperative code.

--- Robert Blumen

Declarative syntax, while more limited, is beneficial for those who prefer a more guided setup 2.

   

Pipeline as Code

Treating Jenkins pipelines as code revolutionizes build and deployment processes. Brent Laster highlights that Jenkins 2 emphasizes storing pipeline configurations as code, integrating them with source control for better version management 3. This approach aligns with DevOps principles, allowing pipelines to be automatically triggered by changes in the Jenkinsfile 3.

Just as you would for source control for a continuous integration. Watch this file. And if something changes in that file, go out and create a job to automatically run my pipeline based on what's in that file.

--- Brent Laster

Hybrid approaches, combining declarative and scripted syntax, offer flexibility, enabling developers to leverage the strengths of both methods 4.

   

Error Management

Managing errors in Jenkins pipelines is crucial for maintaining reliability. Brent Laster discusses the importance of handling exceptions, especially in scripted syntax, where developers must use constructs like try-catch to manage failures 5. Declarative syntax simplifies this with post-section directives that ensure certain actions are executed regardless of pipeline success 5.

The idea with Jenkins two again is that the system is supposed to be reentrant or I jobs should be reentrant, and that's dependent upon the plugins ability again to serialize their data out to disk and be able to pick up then if something happens on the master side.

--- Brent Laster

Scripted syntax solutions, such as retry mechanisms, help in gracefully handling errors without aborting the entire pipeline 6.