I'm using CM 5.4.3 on my cluster which is also running Impala. I tried setting up Hive following the documentation here but it doesn't seem to be the recommended way anymore. Cloudera recommends to leverage Impala to enable stats. I want to use the additional functionalities that come with enabling Hive Stats (vectorization etc.).
The other way is just to enable hive.stats.autogather in the Hive Safety Valve. Will only doing this suffice?