Re: How to increase shared mem for PostgreSQL on FreeBSD

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, 14 Dec 2001, Joe Koenig wrote:

> The situation I'm dealing with is not ideal by any means. All of the
> data is on a different server, and needs to be moved over to mine
> nightly. I guess there is a chance I could have a script connect to the
> other DB, export all of the data in a tab or CSV format, FTP the info
> over to my server and then use COPY to import all of the data. For some
> reason I'm more comfortable with the way it is now than with trying to
> transfer such a large file. I believe I would need to transfer at least
> 100MB of files and then run the copy. Since the FTP transfer would be
> involved, I can't see that saving me any time. I'll look into it though.
> Most of the data being transferred already exists on the current server,
> but the DB on the other side provides no way for me to get only the new
> information. No dates the record was created or updated or anything.
> Like I said, the situation is not at all ideal. By the way, the machine
> has 3 18GB Ultra3 SCSI drives. A couple people have mentioned to drop
> the indexes during the inserts, which I was previously doing. The
> largest (bytes per row, 222,000 rows) table averages around 400 bytes
> per row. So If my math is right, I'm looking at close to 80K/sec. Maybe
> I'm really doing better than I think. Once again, thanks to everyone who
> has replied with comments and advice. I really appreciate it.
>

Would piping the output of pg_dump, through something like netcat, into
psql work? I've never tried it, but I imagine it would work for simple
one-way mirroring....

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iEYEARECAAYFAjwaYGsACgkQwtU6L/A4vVDNYgCfSvwGw8JkAJg0ang1Tvy1ojld
xUQAnijJJ/GV6LraO472IRm3Qwb0ylo/
=13oc
-----END PGP SIGNATURE-----