The standard use case for comprehensive is after you've stood up your cluster and loaded a reasonably representative amount of data.
The standard use case for incremental is you have a query (or two) that you must optimize.
The best things I've learned about the DBD are:
You MUST run it - it's that important
You MUST supply it / point it at as much of your data as possible
You MUST give it a good set of representative queries
It's NOT one and done. In other words you'll want to run this / re-visit it over time. As you continue to load data your schemas will grow and the underlying data changes. As you add users they will create new queries and do things that will most likely require an updated database design..
Comments
Hi. I know I've mentioned this before, and our Vertica Essentials training covers this topic nicely. https://inter.viewcentral.com/events/cust/catalog.aspx?cid=arcsight&pid=1&lid=2&bu_id=6
The standard use case for comprehensive is after you've stood up your cluster and loaded a reasonably representative amount of data.
The standard use case for incremental is you have a query (or two) that you must optimize.
The best things I've learned about the DBD are: