Re: speedup tidbitmap patch: cache page

On 2015-01-16 12:15:35 -0500, Tom Lane wrote:
> Andres Freund <andres(at)2ndquadrant(dot)com> writes:
> > On 2014-12-25 01:26:53 +1300, David Rowley wrote:
> >> So I think v3 is the one to go with, and I can't see any problems with it,
> >> so I'm marking it as ready for committer.
>
> > And committed.
>
> It strikes me that this patch leaves some lookups on the table,
> specifically that it fails to avoid repeated hash_search lookups
> inside tbm_page_is_lossy() in the situation where we're adding
> new TIDs to an already-lossified page. Is it worth adding a few
> more lines to handle that case as well?

There was a alternative version (v2.3 in 549950FB(dot)2050004(at)sigaev(dot)ru) of
the patch that cached the lossyness of a page, but Teodor/David didn't
find it to be beneficial in their benchmarking.

Teodor's argument was basically that it's completely lost in the noise
due to the much bigger overhead of rechecks.

I thought it'd better to get this improvement committed rather than
waiting for someone to find a even bigger improvement for some case.

Andres Freund

--
Andres Freund http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services