On Thu, Mar 15, 2018 at 9:41 AM, Alvaro Herrera <alvherre@alvh.no-ip.org> wrote:
> Tom Lane wrote:
>> Andrew Dunstan <andrew@dunslane.net> writes:
>> > Basing an MV on pg_class could always be difficult for pg_upgrade. Maybe
>> > that's not a brilliant thing to do in a test (or maybe the test should drop
>> > the MV after it's done).
>>
>> OH. Is that what it's doing? The cause of the failure is immediately
>> obvious then. pg_class now lacks a relhaspkey column, but the matview
>> is still dependent on one being there.
>>
>> I concur that this is not well-considered test design.
>
> It seems very strange to me that the test_ddl_deparse database would be
> used by a pg_upgrade test, but OK. I'll change it.
>
As I mentioned upthread, it's not supposed to be, but was being due to
a typo that I have fixed. You should see this error cleared on the
dashboard in a few minutes.
However, in general the module tries to do a maximal test. That
includes almost all the contrib databases. That approach has helped
reveal problems several times in the past.
cheers
andrew
--
Andrew Dunstan https://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services