Re: Mat Views and Conflicts

On 2/20/24 05:25, Zahir Lalani wrote:
> -From: David G. Johnston <david(dot)g(dot)johnston(at)gmail(dot)com>
> -Sent: Tuesday, February 20, 2024 1:20 PM
> -To: Zahir Lalani <ZahirLalani(at)oliver(dot)agency>
> -Cc: Adrian Klaver <adrian(dot)klaver(at)aklaver(dot)com>; pgsql-general(at)lists(dot)postgresql(dot)org
> -Subject: Re: Mat Views and Conflicts
> -
> -
> -On Tuesday, February 20, 2024, Zahir Lalani <mailto:ZahirLalani(at)oliver(dot)agency> wrote:
> -
> -This suggests that not all the MV data is cached and it still queries the source tables in some way?
> -
> -No.  That isn’t how MV work.  If you include an MV relation in your query from clause there is no reference or knowledge as to the underlying query that built the MV physical relation.
> -
> -David J.
>
> Thx David
>
> Cool so that is what my understanding was. But that comes back to my main question - the MV is cached data from the last time refresh was run. If I query the MV and the underlying realtime data changes (i.e. from the source tables not the MV) - should I be getting a replication conflict issue? I would have thought not as the MV data should not have been impacted

Then either something is running REFRESH MATERIALIZED VIEW or some other
process is accessing the underlying tables.

>
> Z

--
Adrian Klaver
adrian(dot)klaver(at)aklaver(dot)com