0
votes

I'm currently trying to optimize my continuous integration process, running on Jenkins, by using Jenkins slaves.
It works very well, but there is one small problem that I can't solve.

On my master Jenkins node, I have only one workspace that is used by all my jobs.
That is good.

On my slave nodes, each job has its own workspace. That is a problem for me, because I have a lot of jobs, and I can't multiply workspaces on my slave nodes that are developers machine.

I search in the node configurations, in Jenkins config files, on the internet, but I found no solution.
Can anyone help me ?

Thank you,

Seb

1

1 Answers

1
votes

If you are using nodes to build your job, it's mandatory to have a local workspace (on the node).

If you want to limit the disk space usage and to avoid workspace duplication, I propose to use Jenkins labels (= one job running only on one dedicated node).

With this solution, the job workspace will exist only in one place.

You cannot use Jenkins slaves just to do distributed builds (like Incredibuild).