| From: | Florian Pflug <fgp(at)phlo(dot)org> |
|---|---|
| To: | Robert Haas <robertmhaas(at)gmail(dot)com> |
| Cc: | Cédric Villemain <cedric(dot)villemain(dot)debian(at)gmail(dot)com>, Bruce Momjian <bruce(at)momjian(dot)us>, Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov>, Simon Riggs <simon(at)2ndquadrant(dot)com>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Subject: | Re: the big picture for index-only scans |
| Date: | 2011-06-19 23:59:44 |
| Message-ID: | 1193BD0C-D495-491F-A519-84319385F012@phlo.org |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Jun19, 2011, at 23:16 , Robert Haas wrote:
> On Sun, Jun 19, 2011 at 5:10 PM, Florian Pflug <fgp(at)phlo(dot)org> wrote:
>> On Jun19, 2011, at 20:40 , Robert Haas wrote:
>>> 2. Since VACUUM and ANALYZE often run together, we will be estimating
>>> the percentage of rows on all-visible pages just at the time when that
>>> percentage is highest. This is not exactly wonderful, either...
>>
>> Hm, doesn't autovacuum run ANALYZE quite a bit more frequently than
>> VACUUM by default?
>
> The autoanalyze threshold is, by default, 10%; and the autovacuum
> threshold, 20%.
Hm, so you could ignore (or rather dampen) the results of
VACUUM+ANALYZE and rely on the ANALYZE-only runs to keep
the estimate correct. Still doesn't sound that bad...
best regards,
Florian Pflug
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Greg Smith | 2011-06-20 00:30:24 | Re: [WIP] cache estimates, cache access cost |
| Previous Message | Florian Pflug | 2011-06-19 23:56:45 | Re: Adding a distinct "pattern" type to resolve the "~" commutator stalemate |