Doug McNaught <[EMAIL PROTECTED]> writes:
> Short answer: MVCC tuple visibility status isn't (and can't be) stored
> in the index.
Well the "can't" part is false or at least unproven. From prior discussion the
only thing that would be technically challenging would be avoiding deadlocks.
The main issue raised is that storing the visibility information in the index
would incur other performance costs. It would increase the storage size of the
index and it would dramatically increase the i/o needed to maintain indexes
for updates and deletes.
So it becomes a question of performance trade-offs, as many things do.
It would be an experiment that would require a lot of work to even try,
and people are skeptical that it would really have any benefits.
--
greg
---------------------------(end of broadcast)---------------------------
TIP 5: Have you checked our extensive FAQ?
http://www.postgresql.org/docs/faqs/FAQ.html
