Re: Reduce maximum error in tuples estimation after vacuum.

<p dir="ltr">Sorry, I made an mistake.<p dir="ltr">"Kyotaro HORIGUCHI" <<a
href="mailto:horiguchi.kyotaro@lab.ntt.co.jp">horiguchi.kyotaro@lab.ntt.co.jp</a>>:<pdir="ltr">> Overall
algorithmcould be illistrated as below,<br /> ><br /> >  - summing up used bytes, max
offnum(PageGetMaxOffsetNumber),<pdir="ltr">Not "max" offnum, the number of linp's used after page vacuum.<p
dir="ltr">>   maximum free bytes for tuple data , and free bytes after page<br /> >    vacuum through all scanned
pages.<br/> ><br /> >  - summing up free bytes informed by FSM through all skipped<br /> >    pages.<br />
><br/> >  - Calculate mean tuple length from the overall used bytes and<br /> >    sum of max offnums, and
scannedpages.<p dir="ltr">Here also is the same. not sum of max offnum but total of used entrre(linp)s.<p
dir="ltr">> - Guess tuple density in skipped pages using overall free bytes<br /> >    from FSM and the mean
tuplelength calculated above.<br /> ><br /> >  - Finally, feed estimated number of the live tuples BEFORE<br />
>   vacuum into vac_estimate_reltuples.<p dir="ltr">regards,<p dir="ltr">-- <br /> Kyotaro Horiguchi