Re: Declarative partitioning

On 2015-08-20 PM 06:34, Simon Riggs wrote:
> On 20 August 2015 at 10:10, Amit Langote <Langote_Amit_f8(at)lab(dot)ntt(dot)co(dot)jp>
> wrote:
>
>> On 2015-08-20 AM 05:10, Josh Berkus wrote:
>>>> PARTITION BY RANGE ON (columns) INCREMENT BY (INTERVAL '1 month' )
>>>> START WITH value;
>>>
>>> Oh, I like that syntax!
>>>
>>> How would it work if there were multiple columns? Maybe we don't want
>>> to allow that for this form?
>>>
>>
>> Yea, we could simply restrict it to the single column case, which does not
>> sound like a major restriction.
>>
>
> PARTITION BY ...
> SUBPARTITION BY ...
>
> We should plan for that in the way we develop the internals, but full
> support can wait until later patches.
>

At the moment, a form of SUBPARTITION BY is to allow PARTITION BY in a
partition definition. But I can see that may not be what people would expect.

> My view has long been that the internals are they aspect here, not the
> syntax. We need to be able to have a very fast partition-selection
> mechanism that can be used in the planner or executor for each tuple.
> Working backwards, we need a relcache representation that allows that, and
> a catalog representation that allows that and syntax to match.
>

Agreed.

Thanks,
Amit