Re: Partial indexes instead of partitions
| От | David Wilson |
|---|---|
| Тема | Re: Partial indexes instead of partitions |
| Дата | |
| Msg-id | AANLkTikU0uEN15wpHlEY-73TWzaA0aWyRlOA1C3mCz87@mail.gmail.com обсуждение исходный текст |
| Ответ на | Re: Partial indexes instead of partitions (Leonardo F <m_lists@yahoo.it>) |
| Ответы |
Re: Partial indexes instead of partitions
Re: Partial indexes instead of partitions |
| Список | pgsql-general |
On Mon, Jun 14, 2010 at 5:24 AM, Leonardo F <m_lists@yahoo.it> wrote:
Do you have any empirical evidence for this being a real problem, or are you simply guessing? I have tables with 500m+ rows, on commodity hardware (4 SATA disks in raid 10), and inserts to the indexes on those tables remain quite acceptable from a performance standpoint.
> For "inserts" I do not see the reasonno, when the index gets very big inserting random values gets
> why
> it would be better to use index partitioning because AFAIK
> b-tree
> would behave exactly the same in both cases.
very slow.
Do you have any empirical evidence for this being a real problem, or are you simply guessing? I have tables with 500m+ rows, on commodity hardware (4 SATA disks in raid 10), and inserts to the indexes on those tables remain quite acceptable from a performance standpoint.
--
- David T. Wilson
david.t.wilson@gmail.com
В списке pgsql-general по дате отправления: