Re: psql copy errors

Поиск
Список
Период
Сортировка
От Scott Marlowe
Тема Re: psql copy errors
Дата
Msg-id 1119624222.8208.67.camel@state.g2switchworks.com
обсуждение исходный текст
Ответ на Re: psql copy errors  ("李江华" <ljh1469@alibaba-inc.com>)
Ответы Re: psql copy errors  (Alvaro Herrera <alvherre@surnet.cl>)
Список pgsql-admin
On Fri, 2005-06-24 at 00:52, 李江华 wrote:
> David Bear:
>
>         Yes. I agree with you.
>         \copy is really too brittle.
>         I wonder why \copy is not like oracle's sqlldr?
>         I think sqlldr is more powerful. When using sqlldr,we can specify the maximum error records we allow,and we
canalso specify the number we should commit in every transaction. 

While I wouldn't mind have the versatility of sqlldr, I would HATE to
have to deal with such a clunky interface.  The only features of sqlldr
I actually like are the ones that allow rejected records to go into a
separate file.

But like most of oracle's tools, it's awkward to use, and requires a lot
of up front work even if you only want a simple load.  Sure, add some
features to \copy, but don't emulate the WAY those features are
implemented oracle.

>        Another aspect is also important. Oracle has better partition table facilities,it's especially suitable for
largetables,as well as index partition concepts.But PostgreSQL has no such concepts.These are really important for
largedatabase. 

No one would complain if someone stepped up and programmed a decent
table partitioning patch.  I have a feeling it's one of the things on
the TODO list that will happen in the next version or two.


В списке pgsql-admin по дате отправления:

Предыдущее
От: Tim.Shenton@corp.terralycos.com
Дата:
Сообщение: Re: symlinking pg_xlog
Следующее
От: Peter Eisentraut
Дата:
Сообщение: Re: unicode