Received: from localhost (postgresql.org [64.49.215.8]) by postgresql.org (Postfix) with ESMTP id 7BD63476179 for ; Fri, 24 Jan 2003 19:50:36 -0500 (EST) Received: from sss.pgh.pa.us (unknown [192.204.191.242]) by postgresql.org (Postfix) with ESMTP id 003AE475E3E for ; Fri, 24 Jan 2003 19:50:21 -0500 (EST) Received: from sss2.sss.pgh.pa.us (tgl@localhost [127.0.0.1]) by sss.pgh.pa.us (8.12.6/8.12.6) with ESMTP id h0P0oO5u027965; Fri, 24 Jan 2003 19:50:24 -0500 (EST) To: Noah Silverman Cc: pgsql-performance@postgresql.org Subject: Re: Multiple databases one directory In-reply-to: References: Comments: In-reply-to Noah Silverman message dated "Fri, 24 Jan 2003 17:39:42 -0500" Date: Fri, 24 Jan 2003 19:50:24 -0500 Message-ID: <27964.1043455824@sss.pgh.pa.us> From: Tom Lane X-Virus-Scanned: by AMaViS new-20020517 X-Archive-Number: 200301/306 X-Sequence-Number: 953 Noah Silverman writes: > We have a situation where we will be creating two fairly large and > complex databases with many tables (thousands) each. From what I > understand, postgres keeps everything in one big data directory. Yeah. You're kind of at the mercy of the operating system when you do that: if it copes well with big directories, no problem, but if lookups in big directories are slow then you'll take a performance hit. The first thing I'd ask is *why* you think you need thousands of tables. How will you keep track of them? Are there really thousands of different table schemas? Maybe you can combine tables by introducing an extra key column. Perhaps a little bit of rethinking will yield a small design screaming to get out of this big one ... regards, tom lane