From: | "Jim C(dot) Nasby" <jim(at)nasby(dot)net> |
---|---|
To: | ITAGAKI Takahiro <itagaki(dot)takahiro(at)oss(dot)ntt(dot)co(dot)jp> |
Cc: | Heikki Linnakangas <heikki(at)enterprisedb(dot)com>, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: Another idea for dealing with cmin/cmax |
Date: | 2006-09-29 14:09:41 |
Message-ID: | 20060929140941.GB90915@nasby.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Fri, Sep 29, 2006 at 01:15:06PM +0900, ITAGAKI Takahiro wrote:
>
> "Jim C. Nasby" <jim(at)nasby(dot)net> wrote:
>
> > The reason I thought of this is because once the transaction commits, we
> > have no use for the cid info. So we could do something like have
> > bgwriter look for tuples that belong to committed transactions before it
> > writes a page, and strip the cid out of them.
>
> Your concept is just like as the experimental method that I suggested before
> in http://archives.postgresql.org/pgsql-hackers/2005-08/msg01193.php
> We can remove cmin and cmax from commited tuples and xmin from frozen tuples
> and we might save some bytes in tuple headers.
>
> However, I think our next goal to shrink the headers is 16 bytes. The headers
> become 23 bytes using phantom cids and we are limited by alignments, so we will
> have no more advantages unless we delete extra 7 bytes in the headers.
> ...and it seems to be very difficult.
Dumb question... wouldn't getting down to 20 bytes buy us something?
--
Jim Nasby jim(at)nasby(dot)net
EnterpriseDB http://enterprisedb.com 512.569.9461 (cell)
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2006-09-29 14:23:31 | Re: Block B-Tree concept |
Previous Message | Markus Schaber | 2006-09-29 14:03:50 | Re: send()/receive() and on-disk storage |