Problem Definition

Currently when Jira Datacenter has lost global task (see related JRASERVER-47045, JRASERVER-66204), there is no way to remove this task. All nodes still thinks task is running, so this prevents corresponding operations to be done. 

Suggested Solution

Create a plugin and/or health-check which lists running cluster tasks, show possible stuck and allow to remove them.

Workaround

Workaround via ScriptRunner in Script Console, From JRASERVER-66204?focusedCommentId=2280767

1. List all live tasks:

   import com.atlassian.jira.component.ComponentAccessor
   import com.atlassian.jira.task.TaskManagerImpl
    
   def TaskManager = ComponentAccessor.getComponent(TaskManagerImpl)
   def tasks = TaskManager.getLiveTasks()
    
   for(key in tasks){
       log.info 'Taskid: ' + key.getTaskId() + ' Description: ' + key.getDescription() + ' Start Time:' + key.getStartedTimestamp() + ' Username: ' + key.getUserName()
   }
   

2. Find stuck job and copy <Taskid> number
3. Kill such Taskid, replacing xxxxxxx with Taskid number

   import com.atlassian.jira.component.ComponentAccessor
   import com.atlassian.jira.task.TaskManagerImpl
    
   def TaskManager = ComponentAccessor.getComponent(TaskManagerImpl)
   TaskManager.removeTask(xxxxxxx)