Re: [PATCH] Stop ALTER SYSTEM from making bad assumptions

On Wed, Jun 19, 2019 at 10:09 AM Ian Barwick
<ian.barwick@2ndquadrant.com> wrote:
>
> On 6/19/19 12:46 PM, Amit Kapila wrote:
> > On Mon, Jun 17, 2019 at 8:20 PM Ian Barwick <ian.barwick@2ndquadrant.com> wrote:
> >> On 6/15/19 1:08 AM, Stephen Frost wrote:
> >>   > * Amit Kapila (amit.kapila16@gmail.com) wrote:
> >>   >> Right.  I think if possible, it should use existing infrastructure to
> >>   >> write to postgresql.auto.conf rather than inventing a new way to
> >>   >> change it.  Apart from this issue, if we support multiple ways to edit
> >>   >> postgresql.auto.conf, we might end up with more problems like this in
> >>   >> the future where one system is not aware of the way file being edited
> >>   >> by another system.
> >>   >
> >>   > I agere that there should have been some effort put into making the way
> >>   > ALTER SYSTEM is modified be consistent between the backend and utilities
> >>   > like pg_basebackup (which would also help third party tools understand
> >>   > how a non-backend application should be modifying the file).
> >>
> >> Did you mean to say "the way postgresql.auto.conf is modified"?
> >>
> >
> > Yes, that is what we are discussing here.  I think what we can do here
> > is to extract the functionality to set the parameter in .auto.conf
> > from AlterSystemSetConfigFile and expose it via a function that takes
> > (option_name, value) as a parameter.
>
> Yup, I was just considering what's involved there, will reply to another
> mail in the thread on that.
>
> > Then we can expose it via some
> > SQL function like set_auto_config (similar to what we have now for
> > set_config/set_config_by_name).  I think if we have something like
> > that then pg_basebackup or any other utility can use it in a
> > consistent way.
>
> Umm, but the point is here, the server will *not* be running at this point,
> so calling an SQL function is out of the question. And if the server
> is running, then you just execute "ALTER SYSTEM".
>

Sure, SQL function will be a by-product of this.  Can't we expose some
function that can be used by base backup?

-- 
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com