Re: Selectivity of "=" (Re: [HACKERS] Index not used on simple select)
От | Bruce Momjian |
---|---|
Тема | Re: Selectivity of "=" (Re: [HACKERS] Index not used on simple select) |
Дата | |
Msg-id | 199907280010.UAA20267@candle.pha.pa.us обсуждение исходный текст |
Ответ на | Re: Selectivity of "=" (Re: [HACKERS] Index not used on simple select) (Hannu Krosing <hannu@trust.ee>) |
Список | pgsql-hackers |
> Tom Lane wrote: > > > It turns out that the selectivity estimate for an "=" comparison is > just > > the attdisbursion statistic calculated by VACUUM ANALYZE, which can be > > roughly defined as the frequency of the most common value in the column. > > (I took statistics too long ago to recall the exact definition.) > > Anyway, given that the test data Ole sent me contains nearly all '' > > entries, I'd say that the 0.995 value is about right for disbursion. > > > > Indeed, if one were to do a "select * from av_parts where nsn = ''", > > then sequential scan would be the most efficient way to do that. > > The system has no clue that that's not really something you'd do much. > > Does the system currently index NULLs as well ? > > I suspect supporting partial indexes (initially just non-NULLs) would > let us have much better and also use indexes intelligently for > mostly-NULL > columns. > > Perhaps a line like > > * Add partial index support > > would fit in TODO > > ----------------- > Hannu > > Yes, I think we index nulls. What are partial indexes? -- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610) 853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania19026
В списке pgsql-hackers по дате отправления: