Decibel! wrote:
> On Tue, Sep 18, 2007 at 11:32:52AM -0400, Tom Lane wrote:
>>> Another option would be to prune whenever the free space goes
>>> below table fillfactor and hope that users would set fillfactor so that
>>> atleast one updated tuple can fit in the block. I know its not best to
>>> rely on the users though. But it can be good hint.
>> If default fillfactor weren't 100% then this might be good ;-). But
>
> Erik Jones and I were just talking about FILLFACTOR...
>
> Is the plan to keep it at 100% with HOT? ISTM that's not such a great
> idea, since it forces at least the first update (if not many more) to be
> COLD.
I think we should still keep it at 100%. Most tables are not updated,
and a non-100% fillfactor will be waste of space when the extra space is
not needed. Even a table that is updated should reach a steady state
after a few cold updates. Those cold updates will make room on the pages
for future updates, now that we can prune them and leave only dead line
pointers behind.
-- Heikki Linnakangas EnterpriseDB http://www.enterprisedb.com