On 9/23/19 3:56 PM, Corey Taylor wrote:
> On Mon, Sep 23, 2019 at 5:51 PM Adrian Klaver <adrian.klaver@aklaver.com
> <mailto:adrian.klaver@aklaver.com>> wrote:
>
> Usually what is seen here is the opposite, that tables are restored and
> ANALYZE is not run and performance on the subsequent queries is poor
> due
> to lack of current statistics.
>
> What is the restore process?
>
>
> For these specific legacy db tables, they are isolated in a separate
> schema. We then use pg_restore to restore the entire schema.
> Essentially just:
>
> pg_restore -n wss --no-owner
Per my previous post and below, the above does not kick off an ANALYZE:
https://www.postgresql.org/docs/11/app-pgrestore.html
"Once restored, it is wise to run ANALYZE on each restored table so the
optimizer has useful statistics; see Section 24.1.3 and Section 24.1.6
for more information."
So is there some other step in the process that occurs after the restore
and before you run your function?
>
> corey
--
Adrian Klaver
adrian.klaver@aklaver.com