Re: Fix for pg_upgrade and invalid indexes

On 2013-03-29 16:57:06 -0400, Bruce Momjian wrote:
> On Thu, Mar 28, 2013 at 05:27:28PM -0400, Tom Lane wrote:
> > Bruce Momjian <bruce@momjian.us> writes:
> > > Should I just patch pg_upgrade to remove the "indisvalid", skip
> > > "indisvalid" indexes, and backpatch it?  Users should be using the
> > > version of pg_upgrade to match new pg_dump.  Is there any case where
> > > they don't match?  Do I still need to check for "indisready"?
> > 
> > Yeah, if you can just ignore !indisvalid indexes that should work fine.
> > I see no need to look at indisready if you're doing that.
> 
> Attached is a patch that implements the suggested pg_upgrade changes of
> not copying invalid indexes now that pg_dump doesn't dump them.  This
> should be backpatched back to 8.4 to match pg_dump.  It might require
> release note updates;  not sure.  Previously pg_upgrade threw an error
> if invalid indexes exist, but only since February, when we released the
> pg_upgrade fix to do this.  You can see the majority of this patch is
> removing that check.

> +              "RIGHT OUTER JOIN pg_catalog.pg_index i "
> +              "       ON (c.oid = i.indexrelid) "
>                "WHERE relkind IN ('r', 'm', 'i'%s) AND "
> +             /* pg_dump only dumps valid indexes;  testing indisready is
> +              * necessary in 9.2, and harmless in earlier/later versions. */
> +              " i.indisvalid IS DISTINCT FROM false AND "
> +              " i.indisready IS DISTINCT FROM false AND "
>       /* exclude possible orphaned temp tables */
>                "  ((n.nspname !~ '^pg_temp_' AND "
>                "    n.nspname !~ '^pg_toast_temp_' AND "

Those columns cannot be NULL, so using IS DISTINCT FROM seems a bit
clumsy.

Greetings,

Andres Freund

-- Andres Freund                       http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training &
Services