Insufficient key column information for updating or refreshing
Workaround: Ensure your Spot instance limit on your EC2 account is sufficient for the number of instances you request in the specified region.
Workaround: In order to be able to clone such a cluster, update Cloudera Manager to 5.12 or later, and wait for Cloudera Director to refresh its internal cluster model, before enabling HA through the Cloudera Manager wizard.
This specific scenario is likely to happen if newly added instances show up as unhealthy in Cloudera Manager, which can cause parcel distribution errors.
Workaround: The best course of action is to try to figure out why the newly-added instance(s) comes up as unhealthy.
Workaround: Follow the manual migration workflow to transfer roles to the new instances. The cluster refresh process fails if an instance group does not contain the requested number of instances.
Cloudera Bug: CLOUD-7036 The database models for clusters with a heterogeneous instance group are not properly updated during upgrade.