> I reran the tests with hashtext(), and created a SUMMARY.HTML chart:
>
> http://momjian.us/expire/TOAST/
>
> What you will see is that pushing TEXT to a TOAST column
> allows quick access to non-TOAST values and single-row TOAST
> values, but accessing all TOAST columns is slower than
> accessing them in the heap, by a factor of 3-18x.
Um, my understanding for this measurement would be to tweak
TOAST_TUPLE_THRESHOLD, and use a fixed TOAST_MAX_CHUNK_SIZE ?
Why would you want to increase the number of toast tuples required for
one column value ?
My expectation would be, that we want to allow a toast tuple to fill a
whole page (TOAST_TUPLES_PER_PAGE = 1),
but keep or reduce the threshold. Maybe we need other logic to find and
use free space for that, though (8.4 material).
Andreas