Re: MAIN, Uncompressed?

On 29 August 2017 at 07:58, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
> On 26 August 2017 at 05:40, Mark Kirkwood <mark(dot)kirkwood(at)catalyst(dot)net(dot)nz> wrote:
>> On 26/08/17 12:18, Simon Riggs wrote:
>>
>>> On 25 August 2017 at 20:53, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>>>>
>>>> Greg Stark <stark(at)mit(dot)edu> writes:
>>>>>
>>>>> I think this is a particularly old piece of code and we're lucky the
>>>>> default heuristics have served well for all this time because I doubt
>>>>> many people fiddle with these storage attributes. The time may have
>>>>> come to come up with a better UI for the storage attributes because
>>>>> people are doing new things (like json) and wanting more control over
>>>>> this heuristic.
>>>>
>>>> Yeah, I could get behind a basic rethinking of the tuptoaster control
>>>> knobs. I'm just not in love with Simon's specific proposal, especially
>>>> not if we can't think of a better name for it than "MAINU".
>>>
>>> Extended/External would be just fine if you could set the toast
>>> target, so I think a better suggestion would be to make "toast_target"
>>> a per-attribute option .
>>>
>>
>> +1, have thought about this myself previously....thank you for bringing it
>> up!
>
> OK, so table-level option for "toast_tuple_target", not attribute-level option
>
> The attached patch and test shows this concept is useful and doesn't
> affect existing data.
>
> For 4x 4000 byte rows:
> * by default we use 1 heap block and 3 toast blocks
> * toast_tuple_target=4080 uses 2 heap blocks and 0 toast blocks

New patch, v2, since one line in the docs failed to apply because of
recent changes.

--
Simon Riggs http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services