I am using Hortonworks Data cloud in AWS. When my cluster is not in use I usually stop the nodes. When I start the nodes again the master node goes to the unhealthy state. And then the issue with all the other services started like Pig stops execution, Hive server went down etc.
What can be the reason behind this issue ?
Since HDCloud for AWS was designed to be used for ephemeral clusters, not for long-running clusters, stopping and starting clusters is not supported:
You should use Cloudbreak instead of HDCloud on AWS if you need to use this stop/restart capability:
Cloudbreak 2.4.1 docs: https://docs.hortonworks.com/HDPDocuments/Cloudbreak/Cloudbreak-2.4.1/content/index.html