Thread

Re: Make TID Scans recalculate the TIDs less often

Andrey Borodin <x4mmm@yandex-team.ru> — 2025-09-17T10:13:00Z

> On 17 Sep 2025, at 14:51, David Rowley <dgrowleyml@gmail.com> wrote:
> 
> What you'd need for this to ever be slow is lots of rescans, so
> something like TID Scan on the inside of a nested loop. If you look at
> ExecReScanTidScan() you'll see it pfrees the tss_TidList, which
> requires that the list gets built all over again on the next call to
> TidNext(). It's the call to TidListEval() that is potentially
> expensive due to the expression evaluation, memory allocation, sorting
> and distinct work.

Occasionally (when dealing with corruption) I do stuff like

begin;
update public.tablename set description = description where ctid in (select ('('||b.blkno::text||','||(x::text)||')')::tid from generate_series(1,300) x, blocks b);

in some forms they are actually joins. Also, pageinspecting things out is always a join (CTAS a copy of table rows that have particular infomask bits). But, fortunately, it's not that frequent case. It's always "plumbing", not a "regular database usage".


Best regards, Andrey Borodin.