Re: Open issues for HOT patch - Mailing list pgsql-hackers

From Pavan Deolasee
Subject Re: Open issues for HOT patch
Date
Msg-id 2e78013d0709180858ofa8e16er954602755acd06b9@mail.gmail.com
Whole thread Raw
In response to Re: Open issues for HOT patch  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers


On 9/18/07, Tom Lane <tgl@sss.pgh.pa.us> wrote:

> Another option would be to prune whenever the free space goes
> below table fillfactor

If default fillfactor weren't 100% then this might be good ;-).  But
we could use max(1-fillfactor, BLCKSZ/8) or some such.

> Yet another option would be to set a hint on the page whenever we
> fail to do HOT update

This would be a good idea independent of anything else, I think.



Or may be a combination of the above two would work well. If the user
has not specified any fillfactor (and so table is using default 100), the
first update is most likely a COLD update and we set the hint bit. The
retired tuple is then pruned before the next update comes and we shall
do HOT update. OTOH if the use has specified a fillfactor less than
to avoid any COLD update, the first update will be HOT. At that point,
the used space would go above fillfactor and hence we shall prune
before the next update comes making room for next HOT update.

Thanks,
Pavan

--
Pavan Deolasee
EnterpriseDB     http://www.enterprisedb.com

pgsql-hackers by date:

Previous
From: Greg Smith
Date:
Subject: Re: Just-in-time Background Writer Patch+Test Results
Next
From: Tom Lane
Date:
Subject: Re: Open issues for HOT patch