> On Sat, Jun 8, 2019 at 5:30 PM Andres Freund <andres@anarazel.de> wrote:
>
> On 2019-06-08 16:03:09 +0200, Dmitry Dolgov wrote:
> > > On Thu, Jun 6, 2019 at 9:06 AM Michael Paquier <michael@paquier.xyz> wrote:
> > > The table AM lookup happens only when creating a table, so we could just get
> > > a failure when attempting to create a table with this incorrect value.
> >
> > is correct, but doesn't it leave the room for some problems in the future with
> > a wrong assumptions about correctness of default_table_access_method?
>
> What do you mean by that?
I didn't have any particular problem in mind, just an abstract and probably
wrong observation. One more observation is that this
> Every single use of default_table_access_method (and similarly
> default_tablespace) has to check the value, because it could be outdated /
> not checked due to wrong context.
for default_tablespace clearly expressed in GetDefaultTablespace function (if
you see something like that, obviously you better use it), but there is nothing
like similar for default_table_access_method so one have to keep it in mind
(although of course it's not a problem so far, since it's being used in only
one place).