Sounds like one of the services is down. Checking their health and restarting them is very easy from the Cloudera Manager homepage if you've run the script to migrate to that kind of deployment. If you haven't the way to do this with Linux's service management is to run 'sudo service [SERVICE_NAME] status' to see if it's running, and 'sudo [SERVICE_NAME] restart' if it's not. The services I'd check in this case are zookeeper-server, hive-server2 and hive-metastore. If any of those aren't running, restart them. If you continue to have problems with those services dying, I'd check /var/log/zookeeper/ and /var/log/hive/ for error messages.