Hi everybody in this video I'm going to demo how to dynamically generate CI configurations for parent-child pipelines. A typical scenario would be when we have a parent pipeline that at runtime generates some configurations, and then we take these configurations that are being created as artifacts as we're gonna see and we feed them to child pipelines, which will run as independent pipelines, so we're gonna start from typical clubs here. Yama file in this case I have two stages: one called set up and triggers in the set up config stage.

We have generate config job that is composed of the generate config script, which I'm going to show it's a simple Ruby script that generates two files called service. A conflict and service be convict in service, a config we're going to have two simple jobs: service, a1 and service a2, and in the service B pipeline. We're gonna have service b1 and service bill too. So I'm gonna see that this files actually don't exist in the repository.

So we only have a red meat is generated, convict and ticket lab CI young and after that running the after running, to generate config scripts. We are going to double check that with its status, that new files that they act and the local directory and then we're gonna cut the two fastest to show the the content of these files.

We are setting these files to be as artifacts to will. They will be uploaded and and then what we could do in the next stage, the trigger stage. We have two separate trigger jobs, one the triggers service, a pipeline and another one trigger service deeper pipeline. These to use the new trigger include syntax, which will generate a child pipeline, but the difference is that we are going to specify the young the input llamo file to be provided by artifact from our burst job, which is they generate config.

And so, let's see how what this looks like one. We run a pipeline, so I have here. The pipeline's, I'm gonna generate run a new one and I'm going to specify the right branch.

So what are we gonna see here? We have so I'm gonna refresh this just to see the progresses immediately. So we have the generate conflict that run and provided this the input to configuration for us for trigger service a trigger. So this will be a. We can see this already running as child pipelines leaves already passed, and this is so. What are we going to see?

Is that if we have a look at the generate config a job, we have the we run the generate config we have the ticket status, so that shows that needs to new files were generated. The content of this two files and there Pingdom uploaded as refresh we actually see that everything escalated.