CICD deployment best practices

Pega-as-a-Service
, , ,
1

Hi Team,

We have multiple built-on applications. We are planning for CICD pipeline implementation. Which one is the best practice?

  1. Deploy the entire application after each branch merge. If we do so, assume we have a different set of data instance values for the Dev and Prod environment. In this case, how can we restrict these data instance updates?

If it has to very manually, it will be easier to move code manually.

  1. Deploy only delta RSV and the data instances
2

@SureshP16655111 If the entire application size is getting too big, you could separate your pipeline per application. For example, you could have a separate pipeline for your enterprise layer application and another pipeline for your implementation layer application. I’ve seen some projects having a separate pipeline per component application. In my project, we have different dedicated pipelines for BIX, SSO, Questionnaire, etc.

On your other question regarding a data instance update during deployment, you may want to review this article - Pegasystems Documentation. In my current project, we use “Override” by default, which can be a more aggressive approach but reduce manual updates; any update in DEV will automatically override in higher envs during deployment. But sometimes clients may want to choose “Skip” and update data instances manually in different environments.

FYI - we use DM 5.5.6 and Platform 8.8.4.