I have a Jenkins in openshift using the kubernetes plugin i am trying to spin slaves on demand.

The plugin does spin up a slave pod however the above behavior happens: the top slave gets connected and in the logs i see the following

"Waiting for agent to connect (0/100):"

Then the second slave connects via JNLP4 and my job runs on that slave but because kubernetes plugin is still waiting for the agent from the first slave it kills my pod. I am not sure whats going on.

In the console I can see that the two slaves connected from the UI are one of the same POD in openshift.