After checking the status of the underlying processes(the process for docker worker & the process which gracefully stopped docker), it was clear that all processes had been stopped. There were not any processes related to docker etc. that were running in the relevant host. Hence the "busy" state of the docker worker was false. I manually updated the state of the service in the underlying cloudera manager database and set it to "stopped" to reflect the actual status of the service similarly to the other services. After I applied that, cloudera manager reflected properly the docker daemon as stopped. Then, I was able to start the docker daemon and the whole CDSW service without any issues. The CDSW service is now fully operational and without any issues. Apparently, some hickup or glitch had prevented the update of the docker daemon state in the database that cloudera manager uses, after the daemon had been stopped. Please, let me know if that is a known bug and what may have caused the issue. Would it make sense to have a reset option for the state of the service in the UI?
... View more
