Re: Multiple FPI_FOR_HINT for the same block during killing btree index items

On Thu, Apr 9, 2020 at 10:08 PM Peter Geoghegan <pg(at)bowt(dot)ie> wrote:
>
> On Thu, Apr 9, 2020 at 6:47 PM James Coleman <jtc331(at)gmail(dot)com> wrote:
> > I believe the write pattern to this table likely looks like:
> > - INSERT
> > - UPDATE
> > - DELETE
> > for every row. But tomorrow I can do some more digging if needed.
>
> The pg_stats.null_frac for the column/index might be interesting here. I
> believe that Active Record will sometimes generate created_at columns
> that sometimes end up containing NULL values. Not sure why.

null_frac is 0 for created_at (what I expected). Also (under current
data) all created_at values are unique except a single row duplicate.

That being said, remember the write pattern above: every row gets
deleted eventually, so there'd be a lots of dead tuples overall.

James