From owner-freebsd-arch Tue Nov 27 11:15:17 2001 Delivered-To: freebsd-arch@freebsd.org Received: from apollo.backplane.com (apollo.backplane.com [216.240.41.2]) by hub.freebsd.org (Postfix) with ESMTP id 4487E37B41C for ; Tue, 27 Nov 2001 11:15:15 -0800 (PST) Received: (from dillon@localhost) by apollo.backplane.com (8.11.6/8.9.1) id fARJEhn93832; Tue, 27 Nov 2001 11:14:43 -0800 (PST) (envelope-from dillon) Date: Tue, 27 Nov 2001 11:14:43 -0800 (PST) From: Matthew Dillon Message-Id: <200111271914.fARJEhn93832@apollo.backplane.com> To: Sheldon Hearn Cc: David Greenman , Kirk McKusick , freebsd-arch@FreeBSD.ORG Subject: Re: Using a larger block size on large filesystems References: <82580.1006875548@axl.seasidesoftware.co.za> Sender: owner-freebsd-arch@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG :I'm almost finished with my batch of postmark benchmarks and will try :your program next. : :Postmark seems to reflect MTA-like access patterns very accurately. :Postmark shows my Mylex eXtremeRAID + 15Krpm storage as about 3 times :faster than my Compaq SmartArray + 10Krpm storage, and Exim shows a :corresponding performance increase approaching a factor of 3. : :Can we rely on your program to be similarly reflective of a RDBMS-like :access patterns? : :Ciao, :Sheldon. It should be reflective of the absolute worst case type of database access - a random seek/read. This is quite typical of what databases handling large numbers of independant transactions (think VISA) have to deal with. It is not reflective of a general purpose RDBMS, since those rely heavily on caching (and would look more like an MTA access pattern). -Matt Matthew Dillon To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message