Re: [bug] Table not have typarray when created by single user mode

Wenjing Zeng <wjzeng2012@gmail.com>

From: wenjing <wjzeng2012@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: pgsql-bugs@lists.postgresql.org, tushar <tushar.ahuja@enterprisedb.com>
Date: 2020-04-14T03:16:36Z
Lists: pgsql-bugs, pgsql-hackers

> 2020年4月13日 下午10:51,Tom Lane <tgl@sss.pgh.pa.us> 写道:
> 
> wenjing <wjzeng2012@gmail.com> writes:
>> Use single user mode (t_create_by_standalone) typarray = 0, but use psql t_create_by_psql typarray has oid.
> 
> That's because of this in heap_create_with_catalog:
> 
>    /*
>     * Decide whether to create an array type over the relation's rowtype. We
>     * do not create any array types for system catalogs (ie, those made
>     * during initdb). We do not create them where the use of a relation as
>     * such is an implementation detail: toast tables, sequences and indexes.
>     */
>    if (IsUnderPostmaster && (relkind == RELKIND_RELATION ||
>                              relkind == RELKIND_VIEW ||
>                              relkind == RELKIND_MATVIEW ||
>                              relkind == RELKIND_FOREIGN_TABLE ||
>                              relkind == RELKIND_COMPOSITE_TYPE ||
>                              relkind == RELKIND_PARTITIONED_TABLE))
>        new_array_oid = AssignTypeArrayOid();
> 
> Admittedly, "!IsUnderPostmaster" is not exactly the same thing as "running
> during initdb", but I do not consider this a bug.  You generally should
> not be using single-user mode for anything except disaster recovery.
> 
> 			regards, tom lane


Thanks for explain. I can understand your point.

However, if such a table exists, an error with pg_upgrade is further raised

./initdb -k -D datanew
./pg_upgrade -d data -d datanew - b. -b.

Restoring database schemas in the new cluster
  postgres                                                  
*failure*

Consult the last few lines of "pg_upgrade_dump_13580.log" for
the probable cause of the failure.
Failure, exiting

pg_restore: from TOC entry 200; 1259 13581 TABLE t_create_by_standalone wenjing
pg_restore: error: could not execute query: ERROR:  pg_type array OID value not set when in binary upgrade mode
Command was:
-- For binary upgrade, must preserve pg_type oid
SELECT pg_catalog.binary_upgrade_set_next_pg_type_oid('13582'::pg_catalog.oid);

I wonder if there are any restrictions that need to be put in somewhere?



Wenjing




Commits

  1. Don't create pg_type entries for sequences or toast tables.

  2. Create composite array types for initdb-created relations.