Under certain conditions:
- multi-node deployment
- LB in front
- task or report created and executed by two subsequent calls to LB, reaching two different back-end nodes (typically from Console)
an error is raised:
org.quartz.SchedulerException: Job instantiation failed [See nested exception: org.springframework.beans.factory.NoSuchBeanDefinitionException: No bean named 'taskJob379a7e29-e7cf-4373-9a7e-29e7cfa373b6' available]
This happens because AutowiringSpringBeanJobFactory is expecting a singleton bean for each task or report to run: such a bean is registered upon task or report create, but remains in memory only for the node on which create has happened.