Re: faster inserting from dump?
| От | Gregory Wood |
|---|---|
| Тема | Re: faster inserting from dump? |
| Дата | |
| Msg-id | 000501c0bdd8$10b1fea0$7889ffcc@comstock.com обсуждение исходный текст |
| Ответ на | faster inserting from dump? ("chris markiewicz" <cmarkiew@commnav.com>) |
| Список | pgsql-general |
> how can i increase the speed of building a database from a dump? there are > currently many tens-of-thousands of insert statements...i assume that a > commit is run after each statement? Correct... every statement is wrapped in an implicit transaction if an explicit one is not provided. > i am guessing that it would be faster > if i commit after every hundred or thousand? Correct again. > is there an easy way to tell psql to do this? Just use BEGIN; and COMMIT; to denote a transaction block. That's it! You might also want to research the COPY command for doing bulks imports. I personally use INSERTs because I need to have my importing error checked, but if you can guarantee the integrity of the import, COPY is your best bet. Greg
В списке pgsql-general по дате отправления: