Re: pg_restore -L reordering of the statements does not work

Thanks David and Tom. Yes, before restoring we are making all the roles
member of <<user1>> but the grant is required since the scenario becomes
very complicated.
Example, following are the steps -
* <<user1>> member of <<user2>>,<<user3>>,<<user4>>
* Restoring the dump file using list file with role <<user1>>
* alter schema schema1 owner to user2
* create table schema1.table1
* create table schema1.table2
* grant all on schema1 to user3
* alter table schema1.table1 to user2
* alter table schema1.table2 to user3

In the above scenario the grant statement for the schema to user3 has to be
provided and is it possible via list file?

Regards,
Aditya D

On Sun, 29 Oct 2023 at 07:19, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:

> "David G. Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com> writes:
> > On Sat, Oct 28, 2023, 18:37 Aditya D <dsaditya91(at)gmail(dot)com> wrote:
> >> PS: Don't have super user privileges on the target.
>
> > Why can you not create a role on the target that is a member of every
> other
> > role and thus can create and reassign delegated permissions and ownership
> > to any of them. In short, a role that behaves as superuser within this
> > limited context even if they are not a true superuser.
>
> That's going to be essential in any case if the restoring user is
> to be able to successfully execute the ALTER OWNER commands in the
> script. You can't give away ownership to a role you don't have
> the privileges of.
>
> regards, tom lane
>