Re: pg_upgrade and relkind filtering - Mailing list pgsql-hackers

On Mon, Dec 5, 2011 at 5:06 PM, Bruce Momjian <bruce@momjian.us> wrote:
> Pg_upgrade has the following check to make sure the cluster is safe for
> upgrading:
>
>        res = executeQueryOrDie(conn,
>                                "SELECT n.nspname, c.relname, a.attname
> "
>                                "FROM   pg_catalog.pg_class c, "
>                                "       pg_catalog.pg_namespace n, "
>                                "       pg_catalog.pg_attribute a "
>                                "WHERE  c.oid = a.attrelid AND "
>                                "       NOT a.attisdropped AND "
>                                "       a.atttypid IN ( "
>          "         'pg_catalog.regproc'::pg_catalog.regtype, "
>          "         'pg_catalog.regprocedure'::pg_catalog.regtype, "
>          "         'pg_catalog.regoper'::pg_catalog.regtype, "
>          "         'pg_catalog.regoperator'::pg_catalog.regtype, "
>        /* regclass.oid is preserved, so 'regclass' is OK */
>        /* regtype.oid is preserved, so 'regtype' is OK */
>          "         'pg_catalog.regconfig'::pg_catalog.regtype, "
>          "         'pg_catalog.regdictionary'::pg_catalog.regtype) AND
> "
>          "     c.relnamespace = n.oid AND "
>          "     n.nspname != 'pg_catalog' AND "
>          "     n.nspname != 'information_schema'");
>
> Based on a report from EnterpriseDB, I noticed that we check all
> pg_class entries, while there are cases where this is unnecessary
> because there is no data behind the entry, e.g. views.  Here are the
> relkinds supported:
>
>        #define       RELKIND_RELATION        'r'       /* ordinary table */
>        #define       RELKIND_INDEX           'i'       /* secondary index */
>        #define       RELKIND_SEQUENCE        'S'       /* sequence object */
>        #define       RELKIND_TOASTVALUE      't'       /* for out-of-line values */
>        #define       RELKIND_VIEW            'v'       /* view */
>        #define       RELKIND_COMPOSITE_TYPE  'c'       /* composite type */
>        #define       RELKIND_FOREIGN_TABLE   'f'       /* foreign table */
>        #define       RELKIND_UNCATALOGED     'u'       /* not yet cataloged */
>
> What types, other than views, can we skip in this query?

It's not obvious to me that anything other than a table or index would matter.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company