From owner-freebsd-questions@FreeBSD.ORG Thu Feb 7 23:19:23 2008 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id CFF1016A417 for ; Thu, 7 Feb 2008 23:19:23 +0000 (UTC) (envelope-from bc979@lafn.org) Received: from zoot.lafn.org (zoot.lafn.ORG [206.117.18.6]) by mx1.freebsd.org (Postfix) with ESMTP id B30A513C448 for ; Thu, 7 Feb 2008 23:19:23 +0000 (UTC) (envelope-from bc979@lafn.org) Received: from [10.0.1.2] (pool-71-116-247-93.lsanca.dsl-w.verizon.net [71.116.247.93]) (authenticated bits=0) by zoot.lafn.org (8.13.6/8.13.4) with ESMTP id m17NJMm2099091 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO) for ; Thu, 7 Feb 2008 15:19:23 -0800 (PST) (envelope-from bc979@lafn.org) Message-Id: <01D2D22B-D22E-408B-A454-980A659777CC@lafn.org> From: Doug Hardie To: freebsd-questions In-Reply-To: <1E7F8770-4E0A-4B6A-842F-9B6D6321E5D7@lafn.org> Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes Content-Transfer-Encoding: 7bit Mime-Version: 1.0 (Apple Message framework v915) Date: Thu, 7 Feb 2008 15:19:22 -0800 References: <200802050127.m151Ro3r087662@banyan.cs.ait.ac.th> <1E7F8770-4E0A-4B6A-842F-9B6D6321E5D7@lafn.org> X-Mailer: Apple Mail (2.915) X-Virus-Scanned: ClamAV 0.88.7/5732/Thu Feb 7 14:45:29 2008 on zoot.lafn.org X-Virus-Status: Clean Subject: Re: Problem with makemap and hash tables X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 07 Feb 2008 23:19:23 -0000 On Feb 4, 2008, at 18:08, Doug Hardie wrote: >> I am just under 300K lines in that map at present. While I have >> always wondered where hash was going to break, I hadn't found the >> limit yet. However, the time to build the hash table is extremely >> large with that many entries. I ran some tests a couple months ago >> using btree for the map type (subject was 'Map size'). I didn't >> test over 300K entries, but it sure built the maps a whole lot >> faster. You might try that and see if it builds or not. I just ran some tests using the BTREE method. I was able to build a database with 1,000,000 entries. It builds successfully and I can read all the entries in it. I believe that BTREE must also have a limit, but haven't found it yet. I doubt I will need to past a million records though so its now academic.