Re: PATCH: logical_work_mem and logical streaming of large in-progress transactions

On Tue, May 5, 2020 at 10:31 AM Dilip Kumar <dilipbalaut(at)gmail(dot)com> wrote:
>
> On Tue, May 5, 2020 at 10:25 AM Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
> >
> > On Tue, May 5, 2020 at 9:27 AM Dilip Kumar <dilipbalaut(at)gmail(dot)com> wrote:
> > >
> > > On Mon, May 4, 2020 at 5:16 PM Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
> > > >
> > > > On Fri, May 1, 2020 at 8:41 PM Dilip Kumar <dilipbalaut(at)gmail(dot)com> wrote:
> > > > >
> > >
> > > > 5. Shouldn't we add a check in table_scan_sample_next_block and
> > > > table_scan_sample_next_tuple APIs as well?
> > >
> > > I am not sure that we need to do that, Because generally, we want to
> > > avoid getting any wrong system table tuple which we can use for taking
> > > some decision or decode tuple. But, I don't think that
> > > table_scan_sample falls under that category.
> > >
> >
> > Hmm, I am asking a check similar to what you have in function
> > table_scan_bitmap_next_block(), can't we have that one?
>
> Yeah we can put that and there is no harm in that, but my point is
> the table_scan_bitmap_next_block and other functions where I have put
> the check are used for fetching the tuple which can be used for
> decoding tuple or taking some decision, but IMHO,
> table_scan_sample_next_tuple is only used for analyzing the table.
>

These will be used in TABLESAMPLE scan. Try something like "select c1
from t1 TABLESAMPLE BERNOULLI(30);". So, I guess these APIs can also
be used to fetch the tuple.

--
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com