I've since found that removing the node{} wrapping the custom buildDeployMaven
step seems to solve the deadlock.
I'm guessing node{} was allocating an executor for the step, which then
immediately requested another executor for it's internal pipeline{} ?
On Thursday, 17 December 2020 at 15:50:
I'm trying to understand a problem with our Jenkins, where it seems
piplines get deadlocked waiting for an executor if we run many of them
simultaneously.
Jenkins 2.271 on Linux x64. We have one master node configured with 4
executors. Running a job singly shows it taking 2 slots in the Build
