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

On Thu, Apr 9, 2020 at 8:32 PM Peter Geoghegan <pg(at)bowt(dot)ie> wrote:
>
> On Thu, Apr 9, 2020 at 5:25 PM Peter Geoghegan <pg(at)bowt(dot)ie> wrote:
> > Was this a low cardinality index in the way I describe? If it was,
> > then we can hope (and maybe even verify) that the Postgres 12 work
> > noticeably ameliorates the problem.
>
> What I really meant was an index where hundreds or even thousands of
> rows for each distinct timestamp value are expected. Not an index
> where almost every row has a distinct timestamp value. Both timestamp
> index patterns are common, obviously.

I'll try to run some numbers tomorrow to confirm, but I believe that
the created_at value is almost (if not completely) unique. So, no,
it's not a low cardinality case like that.

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.

James