Hello community, We have hive setup on a separate EMR cluster with an external metastore on an RDS. Tables are managed with S3 as EMRFs backing. We are using kerberos as our user authentication for all edge nodes and machines on the cluster.
We are noticing that any changes from updates or merge statements to be reversed when a major compaction is triggered (either manual or auto). We can't seem to find any documentation referring to this problem. Wondering if we can get support on this. Thanks!
