Re: Batch Inserts

Hi Doug,

The latter is the case. Only one transaction is done because transactions
cannot be nested and so when you use explicit begin-commit, no autocommit
is done.

/Jørn Cornelius Olsen

Doug Fields <dfields-pg-general(at)pexicom(dot)com>
Sent by: pgsql-general-owner(at)postgresql(dot)org
12-12-2002 00:03

To: "Ricardo Ryoiti S. Junior" <suga(at)netbsd(dot)com(dot)br>
cc: pgsql-general(at)postgresql(dot)org, pgsql-jdbc(at)postgresql(dot)org
Subject: Re: [GENERAL] Batch Inserts

Hi Ricardo, list,

One quick question:

> - If your "data importing" is done via inserts, make sure that
the
>batch uses transactions for each (at least or so) 200 inserts. If you
>don't, each insert will be a transaction, what will slow down you.

I use JDBC and use it with the default "AUTOCOMMIT ON."

Does doing a statement, in one JDBC execution, of the form:

BEGIN WORK; INSERT ... ; INSERT ... ; INSERT ...; COMMIT;

Count as N individual inserts (due to the autocommit setting) or does the
BEGIN WORK;...COMMIT; surrounding it override that setting?

Thanks,

Doug

---------------------------(end of broadcast)---------------------------
TIP 5: Have you checked our extensive FAQ?

http://www.postgresql.org/users-lounge/docs/faq.html