Pavan Deolasee wrote:
> On Thu, Feb 28, 2008 at 3:05 PM, Florian G. Pflug <fgp@phlo.org> wrote:
>> If that works, then we might also want to set the visibility hint bits.
>
> Oh yes. Especially because random time-scattered index scans on
> the table can actually generate multiple writes of a page of a
> read-only table.
I remember that Simon tried to set hint bits as well when he wrote the
"skip WAL on new table" optimization, but there was some issues with it.
I can't remember the details, but I think it was related to commands in
the same transaction seeing the tuples too early. Like triggers, or
portals opened before the COPY.
Hint bits is the critical part of the issue. If you can set the hint
bits, then you can freeze as well, but freezing without setting hint
bits doesn't buy you much.
-- Heikki Linnakangas EnterpriseDB http://www.enterprisedb.com