Actually, might it be an idea to stick the temp tables in their own
"subnode", that won't be loaded unless someone actually intentionally
looks at it?
//mha
Dave Page wrote:
> No, user error.
>
> On Thu, Aug 28, 2008 at 7:53 PM, Devrim GÜNDÜZ <devrim@gunduz.org> wrote:
>> May be a TODO item here.
>>
>> -------- Forwarded Message --------
>>> From: Tom Lane <tgl@sss.pgh.pa.us>
>>> To: Roberts, Jon <Jon.Roberts@asurion.com>
>>> Cc: pgsql-general@postgresql.org
>>> Subject: Re: [GENERAL] temp schemas
>>> Date: Thu, 28 Aug 2008 14:23:10 -0400
>>>
>>> "Roberts, Jon" <Jon.Roberts@asurion.com> writes:
>>>> I am noticing a large number of temp schemas in my database. We use
>>>> temp tables but it doesn't appear that the schemas get dropped for some
>>>> reason.
>>> That's intentional. There doesn't seem a lot of value in dropping a
>>> catalog entry that'll just have to be created again later.
>>>
>>>> This greatly slows down how long it takes pgAdmin to connect
>>>> because it retrieves thousands of pg_temp_% schemas.
>>> Why have you got thousands of them? If you are running with thousands
>>> of active backends, may I suggest a connection pooler?
>>>
>>> (It might be a good idea to fix pgAdmin so it ignores other sessions'
>>> temp schemas, though.)
>>>
>>> regards, tom lane
>>>
>> --
>> Devrim GÜNDÜZ, RHCE
>> devrim~gunduz.org, devrim~PostgreSQL.org, devrim.gunduz~linux.org.tr
>> http://www.gunduz.org
>>
>
>
>