What is the correct way to add dependencies between steps in the CDK Pipelines?

[jaydm26]

Hello,

Consider the following:

step1: pipelines.Step
step2: pipelines.Step
step3: pipelines.Step

What is the correct way to say that step2 depends on step1, and step3 depends on step2. From the documentation, they say that one can use

pipelines.Step.sequence([step1, step2, step3])

or

step2.add_step_dependency(step1)
step3.add_step_dependency(step2)

I wish to know which of the two is the right way to go about defining dependencies between steps and if there are any differences that I may not be aware of.

Link to the Documentation: https://docs.aws.amazon.com/cdk/api/v2/python/aws_cdk.pipelines/Step.html

Thank you in advance.

[VyvaHart]

Hi @jaydm26

These two methods define dependencies between steps in a pipeline, but.. each of them have own pros and cons, which depends on your workflow

• pipelines.Step.sequence([step1, step2, step3]): easier to understand and read. Script looks more 'clean', but it is not good approach for complex dependency graphs (branches or conditional dependencies..), because in this case you are using a linear sequence of steps where each step depends on the previous one

• step2.add_step_dependency(step1) step3.add_step_dependency(step2): more flexible and easy to debug, but.. requires more code to define the dependencies, which can be cumbersome for simple linear sequences

From my point of view better to use linear sequence approach pipelines.Step.sequence([step1, step2, step3]) for more clarity and simplicity. Anyway you could change it if your project will need more granularity or/and flexibility

Good luck bro

[github-actions[bot]]

Hello! Reopening this discussion to make it searchable.