When two workflow jobs try to trigger the same job at the same time, java.hudson.model.Queue#scheduleInternal may detect a duplicate.
When that happens, a second instance of WaitingItem does not get created, meaning one of the workflow runs will have no listeners related to the triggered job.
When the triggered job ends, only one workflow run will be notified and continue, the other one will stay listening forever.
The workaround I've had to apply to deal with this is running the following in scriptler:
|Status||Open [ 1 ]||In Progress [ 3 ]|
|Remote Link||This issue links to "PR 121 (Web Link)" [ 12807 ]|
|Resolution||Fixed [ 1 ]|
|Status||In Progress [ 3 ]||Resolved [ 5 ]|
|Workflow||JNJira [ 162666 ]||JNJira + In-Review [ 197025 ]|
|Component/s||pipeline-general [ 21692 ]|
|Component/s||workflow-plugin [ 18820 ]|