| From: | Heikki Linnakangas <heikki(at)enterprisedb(dot)com> |
|---|---|
| To: | Heikki Linnakangas <heikki(at)enterprisedb(dot)com> |
| Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Bruce Momjian <bruce(at)momjian(dot)us>, "Jim C(dot) Nasby" <jim(at)nasby(dot)net>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Block B-Tree concept |
| Date: | 2006-09-29 09:51:32 |
| Message-ID: | 451CECA4.90300@enterprisedb.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Heikki Linnakangas wrote:
> If we want to keep the property that VACUUM doesn't re-evaluate index
> entries, any proposal that doesn't keep track of every heap tuple
> isn't going to work. I'll try to modify the design I had in mind so
> that it does keep track of every heap tuple in some form.
After some thought:
Imagine a normal B-tree just like what we have now. But when there is
more than one tuple on the same heap page with consecutive index keys,
we represent all of them in a single index tuple that contains the key
of the first one of them, and a (run-length encoded) bitmap of the
OffsetNumbers. We should get most of the space and I/O savings as with
the original proposal, but we can vacuum it without re-evaluating index
expressions.
It does change the format of an index tuple, unlike the original
proposal. That adds some complexity. but it's doable.
--
Heikki Linnakangas
EnterpriseDB http://www.enterprisedb.com
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Martijn van Oosterhout | 2006-09-29 09:52:59 | Re: Another idea for dealing with cmin/cmax |
| Previous Message | Markus Schaber | 2006-09-29 09:21:21 | Re: Faster StrNCpy |