Date: Sun, 7 Jan 2007 20:44:40 -0600 From: Kirk Strauser <kirk@strauser.com> To: freebsd-questions@freebsd.org Subject: Re: Tuning PostgreSQL for bulk imports Message-ID: <200701072044.47418.kirk@strauser.com> In-Reply-To: <enruui$v68$1@sea.gmane.org> References: <200701071301.27423.kirk@strauser.com> <200701071618.11242.kirk@strauser.com> <enruui$v68$1@sea.gmane.org>
next in thread | previous in thread | raw e-mail | index | archive | help
[-- Attachment #1 --] On Sunday 07 January 2007 17:15, Ivan Voras wrote: > What are your current IO rates? Since you have only two drives you might be > restricted by available disk bandwidth... So says gstat. According to top, the import only takes about 40% of one CPU, but both of the drives are saturated. I'm contemplating mounting that filesystem async and letting PostgreSQL decide when to sync it. Although I'm sort of allergic to the idea, this seems like it might be relatively safe since no other process touches that data. Any thoughts? The biggest benefit I'm getting from SMP right now is that I can run the Foxpro->PostgreSQL conversion on one processor while postmaster grinds away on the other. > I've just remembered - do you run VACUUM ANALYZE after your big imports? > You should (to get performance on SELECTs afterwards, but it won't help > the imports themselves). Actually, I've been letting the vacuum daemon make that decision for the last few months with no perceptible ill effects. > Ok. Be careful not to use SELECT ... LIMIT ALL ;) LOL! Duly noted. -- Kirk Strauser [-- Attachment #2 --] -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (FreeBSD) iD8DBQBFobAf5sRg+Y0CpvERAvygAJ0YAZ68Mb4/hW422Gbr1Cs56kOvkQCeO1Qj zva1cztYriLGrR+VWhOZHRY= =JRSv -----END PGP SIGNATURE-----
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200701072044.47418.kirk>