Re: problems maintaining boolean columns in a large table

Поиск
Список
Период
Сортировка
От Ben Campbell
Тема Re: problems maintaining boolean columns in a large table
Дата
Msg-id 4B72C773.9060705@scumways.com
обсуждение исходный текст
Ответ на Re: problems maintaining boolean columns in a large table  (Richard Huxton <dev@archonet.com>)
Ответы Re: problems maintaining boolean columns in a large table  (Filip Rembiałkowski <plk.zuber@gmail.com>)
Список pgsql-general
Richard Huxton wrote:
> On 09/02/10 11:25, Ben Campbell wrote:
[I was talking about moving a "needs_indexing" flag out of a big table
into it's own table]
>> But my gut feeling is that the flag would be better off in it's own
>> table anyway, eg:
>>
>> CREATE TABLE needs_indexing (
>> article_id integer references article(id)
>> );

> That sounds sensible to me

Cool - glad to know I'm not suggesting something totally insane! I never
can quite tell when I'm doing database stuff :-)

> Oh - you might want to consider how/whether to handle multiple entries
> for the same article in your queue.

I settled on:

CREATE TABLE needs_indexing (
   article_id integer REFERENCES article(id) PRIMARY KEY
);

The primary key-ness enforces uniqueness, and any time I want to add an
article to the queue I just make sure I do a DELETE before the INSERT.
Bound to be more efficient ways to do it, but it works.

Thanks,
Ben.



В списке pgsql-general по дате отправления:

Предыдущее
От: Pavel Stehule
Дата:
Сообщение: Re: Orafce concat operator
Следующее
От: Greg Stark
Дата:
Сообщение: Re: Best way to handle multi-billion row read-only table?