| From: | "Zeugswetter Andreas ADI SD" <ZeugswetterA(at)spardat(dot)at> |
|---|---|
| To: | "Pavan Deolasee" <pavan(dot)deolasee(at)gmail(dot)com>, "Simon Riggs" <simon(at)2ndquadrant(dot)com> |
| Cc: | "Heikki Linnakangas" <heikki(at)enterprisedb(dot)com>, "Hannu Krosing" <hannu(at)skype(dot)net>, "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>, <pgsql-hackers(at)postgresql(dot)org>, "Pavan Deolasee" <pavan(dot)deolasee(at)enterprisedb(dot)com>, "Nikhil S" <nikhil(dot)sontakke(at)enterprisedb(dot)com> |
| Subject: | Re: HOT for PostgreSQL 8.3 |
| Date: | 2007-02-22 10:57:58 |
| Message-ID: | E1539E0ED7043848906A8FF995BDA57901CAF63C@m0143.s-mxs.net |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
> > I very much like Hannu's idea, but it does present some issues.
> >
> >
> I too liked Hannu's idea initially, but Tom raised a valid
> concern that it does not address the basic issue of root
> tuples. According to the idea, a DEAD root tuple can be used
> for a subsequent update of the same row.
If you are reusing the existing slot of a root tuple how will that
slot likely have room for an extra pointer and a live tuple ?
If the idea does not cover root reuse we don't need pointers.
Imho we should follow the swing idea. It should even be possible
to point the root to the newest dead tuple during update (if it
was index path), and reuse an older dead slot from the chain.
Then we can limit the chain to number of potentially visible
tuples + root + 2 without vacuum.
Andreas
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Simon Riggs | 2007-02-22 10:59:13 | Re: [previously on HACKERS] "Compacting" a relation |
| Previous Message | Peter Eisentraut | 2007-02-22 10:31:55 | Re: SCMS question |