sapling

One way to look at - You don't need to make a right choice. Make a choice and then make it right.

I was wrong.It isn't column partitioning.

Sorry I am wrong.I saw the PARTITION BY {column name} and lazy me assumed things.

It sounds like this is column level partitioning.Each column or columns (based on partitioning expression) is stored as different subtable (or something similar) on disk.If only few columns are frequently accessed, they can be put on cache/faster disk or other neat optimizations for join processing.

you work in Sweden ?