| Karl Wirth Thanks for your comment. But I can't understand well. I think process is like below,
- perforce create workspace for reading Jenkinsfile(s) and creating pipeline. (In this case, I want to create workspace and read it and delete workspace automatically)
- and then running pipeline of some branch,
- kubernetes create pod for this
- and perforce create workspace written in Jenkinsfile
- and sync file, compile, and delete workspace automatically I want
In this senario, why need master server ? I don't know which action is needed for workspace I just think create workspace and then sync and the delete it. and If so many jobs in Jenkins and scheduling very often, is that correct way ? I don't think so, Jenkins should use broker or build farm replica . |