Our Jenkins server has 100+ jobs. We have been dealing with a severe memory leak after we updated our PTC plugin. After trying many things I noticed that if all configurations have unique scm names that this problem goes away.

To recreate this just create a number of jobs with the same scm config name. force a run of all these jobs at the same time ( you might need to have multiple executors ). you should see that some derby write threads never get cleaned up and memory is leaking.