Received: from sss.sss.pgh.pa.us (sss.pgh.pa.us [206.210.65.6]) by hub.org (8.9.2/8.9.1) with ESMTP id KAA21044 for ; Wed, 3 Mar 1999 10:26:14 -0500 (EST) (envelope-from tgl@sss.pgh.pa.us) Received: from sss.sss.pgh.pa.us (localhost [127.0.0.1]) by sss.sss.pgh.pa.us (8.9.1/8.9.1) with ESMTP id KAA25095 for ; Wed, 3 Mar 1999 10:25:31 -0500 (EST) To: pgsql-sql@hub.org Subject: Re: question about pg_dump In-reply-to: Your message of Tue, 2 Mar 1999 23:00:31 -0500 (EST) <199903030400.XAA44981@hub.org> Date: Wed, 03 Mar 1999 10:25:31 -0500 Message-ID: <25093.920474731@sss.pgh.pa.us> From: Tom Lane Bill Brandt writes: > I'm trying to upgrade from 6.3. In doing this, I performed a pg_dump of > each database (pg_dump -o where the oid's matter). I have one database > (which doesn't use oid info in the tables btw) that is around 40MB of > data. The output file that created looks fine, but when I go to load the > data, it gets a portion of the way into loading and indicates that you > cannot exceed 20000 lines of input on a copy command. Is there a way to > dump the database that it won't try more than 20000 lines at a time? You can just split the command into sections by editing the dump file (assuming you have an editor that will cope ;-)): COPY table FROM stdin; blah blah blah blah \. COPY table FROM stdin; more more more more \. COPY is really APPEND, it doesn't delete any tuples that were there, so this should work. Still, a 20K-line limit sounds awfully brain-dead to me. And a quick look at the source code (both 6.4.2 and current CVS) doesn't show any such check unless I missed something. What version are you using? What's the exact text of the error message? regards, tom lane