Tom Lane wrote:
> This is not only really ugly, but 100% toast-specific. The
> qualified-name approach ("toast.autovacuum_enabled") has at least
> a chance of being good for something else. Or just make it
> toast_autovacuum_enabled and do the translation magic at some low
> level in the stateme
Tom Lane wrote:
> Alvaro Herrera writes:
>> I need some more opinions on whether I should continue working here, or
>> stop and leave it for 8.5.
>
> Bruce and I were just talking yesterday about the need to start closing
> down this commitfest. I'm not sure what the schedule is going to end
> u
Alvaro Herrera writes:
> I need some more opinions on whether I should continue working here, or
> stop and leave it for 8.5.
Bruce and I were just talking yesterday about the need to start closing
down this commitfest. I'm not sure what the schedule is going to end
up being; but if you can't se
Dave Page wrote:
> Are we expecting this patch (or whatever it turns into) to go into
> 8.4? It was marked as WIP when feature freeze started and clearly
> still is quite undefined at this stage.
Right. This is a fair objection. I started just by reviewing the
autovacuum-in-reloptions patch, bu
On Wed, Dec 31, 2008 at 9:45 PM, Tom Lane wrote:
> Alvaro Herrera writes:
>> Peter Eisentraut wrote:
>>> On Sunday 21 December 2008 01:48:42 Alvaro Herrera wrote:
ALTER TABLE foo SET (TOAST autovacuum_enabled = false);
ALTER TABLE foo SET (toast.autovacuum_enabled = false);
ALTER T
Alvaro Herrera writes:
> Peter Eisentraut wrote:
>> On Sunday 21 December 2008 01:48:42 Alvaro Herrera wrote:
>>> ALTER TABLE foo SET (TOAST autovacuum_enabled = false);
>>> ALTER TABLE foo SET (toast.autovacuum_enabled = false);
>>> ALTER TABLE foo TOAST SET (autovacuum_enabled = false);
>>> ALTE
Peter Eisentraut wrote:
> On Sunday 21 December 2008 01:48:42 Alvaro Herrera wrote:
> > ALTER TABLE foo SET (TOAST autovacuum_enabled = false);
> > ALTER TABLE foo SET (toast.autovacuum_enabled = false);
> > ALTER TABLE foo TOAST SET (autovacuum_enabled = false);
> > ALTER TABLE foo SET TOAST (auto
On Sunday 21 December 2008 01:48:42 Alvaro Herrera wrote:
> ALTER TABLE foo SET (TOAST autovacuum_enabled = false);
> ALTER TABLE foo SET (toast.autovacuum_enabled = false);
> ALTER TABLE foo TOAST SET (autovacuum_enabled = false);
> ALTER TABLE foo SET TOAST (autovacuum_enabled = false);
The last
On 12/20/08, Alvaro Herrera wrote:
>
> ALTER TABLE foo SET (TOAST autovacuum_enabled = false);
...
> ALTER TABLE foo SET TOAST (autovacuum_enabled = false);
>
i will be happy with any of this options (actually i prefer the second
one but don't have a strong argument against the first)
--
Atenta
Tom Lane napsal(a):
Zdenek Kotala writes:
Another potential problem with toast setting is that reloption is toastable and
it could generates loops in detoasting pg_class tuples. For example toast chunk
size cannot be implement like reloption (or pg_class should use every time
default values).
Zdenek Kotala writes:
> Another potential problem with toast setting is that reloption is toastable
> and
> it could generates loops in detoasting pg_class tuples. For example toast
> chunk
> size cannot be implement like reloption (or pg_class should use every time
> default values).
Nonsen
Alvaro Herrera napsal(a):
ALTER TABLE foo SET (toast.autovacuum_enabled = false);
+1
Do not forget on toast index as well.
ALTER TABLE foo SET (toast_idx.fillfactor = 50);
Another potential problem with toast setting is that reloption is toastable and
it could generates loops in detoasting
Alvaro Herrera writes:
> Euler Taveira de Oliveira wrote:
>> Yes, please. But i'm afraid it is too 'complicated' to expose
>> 'pg_toast.pg_toast_x' to user (but we can solve it with good
>> documentation).
> Hmm, now that I look at that again, it seems a very bad idea.
Yeah --- whatever sol
Euler Taveira de Oliveira wrote:
> Alvaro Herrera escreveu:
> > I'm wondering if I should just allow all reloptions (including
> > fillfactor) or just the autovacuum ones.
> >
> Yes, please. But i'm afraid it is too 'complicated' to expose
> 'pg_toast.pg_toast_x' to user (but we can solve it w
Alvaro Herrera escreveu:
> I'm wondering if I should just allow all reloptions (including
> fillfactor) or just the autovacuum ones.
>
Yes, please. But i'm afraid it is too 'complicated' to expose
'pg_toast.pg_toast_x' to user (but we can solve it with good
documentation). What about xxx_toast
Right now we don't allow setting reloptions to toast tables:
=# alter table pg_toast.pg_toast_16395 set (fillfactor = 40);
ERROR: "pg_toast_16395" is not a table or index
However this is needed for autovacuum, per previous discussion.
I'm wondering if I should just allow all reloptions (includi
16 matches
Mail list logo
