Re: benchmarking the query planner - Mailing list pgsql-hackers

From Vladimir Sitnikov
Subject Re: benchmarking the query planner
Date
Msg-id 1d709ecc0812111354m2f5a7ba5gebc9e1d50b9de913@mail.gmail.com
Whole thread Raw
In response to Re: benchmarking the query planner  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: benchmarking the query planner  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers

I think that would likely be redundant with the histogram.
I am afraid I do not get you. I mean histograms should be considered when it comes to increasing number of MCV entries (at least for numeric/timestamp values). With histogram lower number of entries could be sufficient to get reasonable accuracy of estimations.
I have no idea how to decide between automatic switch between histograms and MCV. It might sound crazy one could compute both histograms and MCV and use them both (and pick an average of two estimations :) )

Regards,
Vladimir Sitnikov

pgsql-hackers by date:

Previous
From: "Nathan Boley"
Date:
Subject: Re: benchmarking the query planner
Next
From: Simon Riggs
Date:
Subject: Re: benchmarking the query planner