My cluster version is 8.5.2 (self-managed) .
I have a long-term (maybe a few days) service task and the task which mentioned in the exception is after the long-term task.
The error tells you pretty much exactly what you said…
Connection failed after 28800s (8h)
Does the task itself take more than 8h to run? Basically, Camunda was trying to send jobs to the worker, and gave up after 8h. We probably won’t be able to tell you why without a lot more information.
The red arrow points to the long-term task.It periodically checks the specified path and processes the files if they exists.
The tasks in the red box threw the DEADLINE exception.