On Tue, Apr 20, 2010 at 10:06 AM, Scott Marlowe <scott.marlowe@gmail.com> wrote:
> On Tue, Apr 13, 2010 at 12:16 PM, Mitch Collinsworth
> <mitch@ccmr.cornell.edu> wrote:
>>
>> Hi Folks,
>>
>> I thought I understood how this worked, but now it's baffling me.
>> I want to drop a database named coral from my test server and load
>> a fresh dump of it from my production server. This has worked when
>> I've done it before, but now it's giving me fits. Here's an
>> example:
>>
>> coral=# \c postgres
>> You are now connected to database "postgres".
>> postgres=# drop database coral;
>> DROP DATABASE
>> postgres=# create database coral;
>> CREATE DATABASE
>> postgres=# \c coral
>> You are now connected to database "coral".
>> coral=# \dt *.*
>> List of relations
>> Schema | Name | Type | Owner
>> --------------------+--------------------------------+-------+----------
>> accmgr | acct_rate | table | accmgr
>> accmgr | acct_rate_working | table | accmgr
>> accmgr | acct_sum | table | accmgr
>> accmgr | acct_sum_raw | table | accmgr
>
> I'm pretty sure that at some time you accidentally created those
> tables in template1 and now you're getting them every time you create
> a database. Clean out template1 and things should be ok.
Note that if you have thousands of objects in template1, you can drop
it and recreate it from template0 like so:
update pg_database set datistemplate=false where datname='template1';
drop database template1;
create database template1 with template template0;
update pg_database set datistemplate=true where datname='template1';
\c template1
vacuum freeze;
