Re: BUG #13963: pg_clusterupgrade and materialized view

On Thu, Feb 18, 2016 at 07:10:55AM +0000, Miha Vrhovnik wrote:
> On 2016-02-18 03:52, Bruce Momjian wrote:
> >On Tue, Feb 16, 2016 at 09:13:11AM +0000, miha(dot)vrhovnik(at)cordia(dot)si
> >wrote:
> >>The following bug has been logged on the website:
> >>
> >>Bug reference: 13963
> >>Logged by: Miha
> >>Email address: miha(dot)vrhovnik(at)cordia(dot)si
> >>PostgreSQL version: 9.5.1
> >>Operating system: Ubuntu
> >>Description:
> >>
> >>Hey,
> >>
> >>I've just made a pg_clusterupgrade where the one database has a
> >>materialized
> >>view which is on another server, so the pg_fdw is used.. and
> >>this seems that
> >>was not updated during the pg_clusterupgrade
> >>
> >>It would be at least nice if pg_clusterupgrade would print out a
> >>notice if
> >>it cannot refresh the view.
> >>
> >>I haven't tested but maybe the problem is the same if one doesn't use
> >>materialized view via the pg_fdw.
> >
> >Why do you think an upgrade should refresh a materialzied view?
> Because after the upgrade the view was empty. I have a cron job
> which calls refresh materialized view with concurently keyword and
> the error returned was cannot refresh empty view concurrently.

Oh, that is odd. So you had to repopulate the view rather than
refreshing it? I am guessing that pg_dump is causing this, though there
is no mention in the docs about it.

--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://enterprisedb.com

+ As you are, so once was I. As I am, so you will be. +
+ Roman grave inscription +