| From: | Francisco Reyes <lists(at)natserv(dot)com> |
|---|---|
| To: | Matthew Kennedy <mkennedy(at)opushealthcare(dot)com> |
| Cc: | pgsql-general(at)postgresql(dot)org |
| Subject: | Re: copy vs. insert w/ no autocommit |
| Date: | 2001-12-20 00:37:54 |
| Message-ID: | 20011219193418.R5735-100000@zoraida.natserv.net |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
On 18 Dec 2001, Matthew Kennedy wrote:
> I realize a COPY statement is much faster than individual INSERT
> statements.
How? Have you tested it? Or do you mean that from reading people's
comments you believe this to be "common knowledge".
>But how does COPY compare to individual INSERTS within a
>BEGIN/END or with setAutoCommit(false)?
Are we talking "copy" or "\copy"?
I am not sure about \copy, but "copy" as far as I know doesn't update all
the stats/indexes so it should be significantly faster to use a copy.
So far for all my big jobs (anything greater than a few thousand rows) I
always use copy and it is fairly fast.
Have you tried doing some testing? How many indexes do you have on the
table?
> If there's a difference, are we
> talking orders-of-magnitude in difference?
Depending on your disk subsistem I would expect copy should be several
times faster due to index overhead.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | t.ngockhoi | 2001-12-20 08:38:15 | Operation eq with varchar |
| Previous Message | Matt Sauve-Frankel | 2001-12-20 00:28:00 | Re: More on Too Many Open Files... NetBSD |