Re: Non-atomic mview refresh and unique index.

From: Yakov Vasilchenko <>
Date: Tue, 22 Oct 2013 15:19:40 +0300
Message-ID: <>

Hi Mohamed,
This is the very valid points.

However what bugs me is the fact that unusable index seems valid to optimizer to do left-join elimination.
And I cannot find it to be written somewhere that non-atomic refresh makes no promises in case of errors during refresh.

Received on Tue Oct 22 2013 - 14:19:40 CEST

Original text of this message