After running ./update-pipelines.sh install, following pipeline jobs are created in your OpenShift project:

1. waiverdb-dev-pipeline
   This pipeline job should be triggered when a PR is created,updated, or merged, which matches the C3I pre-merge/post-merge workflows. Build parameter WAIVERDB_GIT_REF must be provided for the Git commid ID/branch when triggering a build.
   It will
2. Run Flake8 and Pylint checks
3. Run unit tests
4. Build Docs
5. Publish Docs (only when it's build against the master branch)
6. Build SRPM
7. Build RPM
8. Invoke Rpmlint
9. Build container
10. Run functional tests

11. Push container (only when it's build against the master branch)

12. waiverdb-dev-integration-test
    It runs integration test against the current :latest (dev) image.

13. waiverdb-stage-integration-test
    It runs integration test against the current :stage image.

14. waiverdb-prod-integration-test
    It runs integration test against the current :prod image.

15. waiverdb-integration-test
    It runs integration test against a user-specified image. Build parameter IMAGE must be provided.

16. waiverdb-promoting-to-stage
    It promotes the current :latest image to :stage.

17. waiverdb-promoting-to-prod
    It promotes the current :stage image to :prod.

I am also thinking moving the parameter values to separated files. Any good idea for tracking pipeline job instances and their parameter values?

Nice, I was not aware of this syntax. Though SpellCheck warns: "This default assignment may cause DoS due to globbing. Quote it."

s/Intergration/Integration/
s/intergration/integration/

Updated.

+1

This depends on #223 and #225, right?

Did you consider using make or ansible for the purpose of putting all the things together?

Writing code for this feels a little bit exaggerated.

@csomh Makefile sounds a good choice. I would prefer not using ansible for this purpose because it seems to be too heavy.
I am thinking creating some new directories in openshift/, moving the templates and parameters to corresponding directories, then using a Makefile to batch update them.

For example, the directory structure for pipeline jobs waiverdb-dev-integration-test, waiverdb-stage-integration-test, and waiverdb-prod-integration-test coming from the template waiverdb-integration-test-template.yml:

openshift
├── containers
│   ├── jenkins-slave
│   │   ├── Dockerfile
│   └── waiverdb
│       └── Dockerfile
├── pipelines
│   ├── jobs
│   │   ├── waiverdb-dev-integration-test.env
│   │   └── waiverdb-dev-integration-test.tmpl
│   │   ├── waiverdb-stage-integration-test.env
│   │   └── waiverdb-stage-integration-test.tmpl
│   │   ├── waiverdb-prod-integration-test.env
│   │   └── waiverdb-prod-integration-test.tmpl
│   ├── Makefile
│   └── templates
│       ├── waiverdb-integration-test.Jenkinsfile
│       └── waiverdb-integration-test-template.yml
├── README.md

Parameters for a job are stored in file jobs/<job-name>.env, and the template the job uses is stored in file jobs/<job-name>.tmpl

For waiverdb-stage-integration-test pipeline job, api resources are created by:
oc process -f ./templates/waiverdb-integration-test-template.yml --param-file=./jobs/waiverdb-stage-integration-test.env.

Cool! This seems to be a good direction to follow!

@csomh I've updated https://pagure.io/waiverdb/pull-request/225 to match this new pattern.

Did you forgot to commit some files? The commit now contains only *.tmpl and *.env files and commit message doesn't match.

@lholecek Updated.

You can mention how JOBS_DIR Makefile variable and*.env/*.tmpl files are used.

@lhellebr updated

+1

@csomh Is this OK to merge or do you have any comments?

Commit db3670a2 fixes this pull-request