https://www.postgresql.org/docs/current/static/ddl-partition...
As of now, you still need to create the root partitioned table as one command specifying the partitioning method (list or range), partitioning columns (aka PARTITION BY LIST | RANGE (<columns>)) and then a command for every partition specifying the partition bounds. No triggers or CHECK constraints anymore though. Why that way? Because we then don't have to assume any particular use case, for which to provide a shorthand syntax -- like fixed width/interval range partitions, etc.
That said, having the syntax described at the beginning of the last paragraph in the initial version does not preclude offering a shorthand syntax in later releases, as, and if we figure out that offering some such syntax for more common use cases is useful after all.
Because table partitioning is less general than table inheritance, it
is hoped that it will be easier to reason about properties of
partitions, and therefore that this will serve as a better foundation
for a variety of possible optimizations, including query planner
optimizations.If i got this patch right, each partitioned table will have the same data structure and store whole rows (it's even more restrictive than previous inheritance mechanism that allowed extending by adding additional columns).
Column or expression should only define in which table an inserted row is supposed to be stored. A single row will never been torn apart. Still it look like a foundation that facilitate sharding BigData(Set) between multiple servers when used in conjunction with foreign data. However a lot of performance improvements will still be needed to compete against solid NoSQL projects (in which you really have a BigData use case).
But looking a bit forward, developing performances improvements on top of an ACID compliant distributed database seems less difficult than to develop a NoSQL project for it to become ACID.
But this patch is about horizontal partitions.