From owner-freebsd-questions Wed Dec 17 12:02:55 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id MAA08139 for questions-outgoing; Wed, 17 Dec 1997 12:02:55 -0800 (PST) (envelope-from owner-freebsd-questions) Received: from silvester.zoom.es (silvester.zoom.es [195.76.150.4]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id MAA08117 for ; Wed, 17 Dec 1997 12:02:45 -0800 (PST) (envelope-from amora@zoom.es) Received: from zoom130.zoom.es (zoom130.zoom.es [195.76.150.130]) by silvester.zoom.es (8.7.5/8.7.3) with SMTP id WAA10767; Wed, 17 Dec 1997 22:01:07 +0100 Message-Id: <199712172101.WAA10767@silvester.zoom.es> Comments: Authenticated sender is From: "Jesus A. Mora Marin" To: "John S. Dyson" Date: Wed, 17 Dec 1997 21:01:07 +0100 Subject: Re: MFS related questions Reply-to: amora@zoom.es CC: questions@FreeBSD.ORG Priority: normal In-reply-to: <199712170259.VAA01348@dyson.iquest.net> References: <199712162137.WAA09873@silvester.zoom.es> from "Jesus A. Mora Marin" at "Dec 16, 97 09:38:08 pm" X-mailer: Pegasus Mail for Windows (v2.54) Sender: owner-freebsd-questions@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Thank you, John. > MFS performs very well in FreeBSD. Paging only happens when the > files on the MFS filesystem get too big. There is a signficant > design flaw that the buffer cache is synced into the MFS server, > but that really doesn't have a severe performance hit. Yes, I think this point was remarked in McKusick et al's book (pp 305 & 306). It seems this flaw was a result of the virtual memory system implementation when MFS was originally designed. Well, I have thought further about this subject and guess that the MFS_ROOT specifies the size of the mfs root directory when mounted, but I assume this value can be overridden by actual parameters in the mfs_mount command. One last question: how many mfs filesystems can be mounted at any time? Jesus ------ Jesus A. Mora Marin, MD. (aka EA7HAC, ex-EC7DVE)