From owner-freebsd-hackers Fri Jan 3 02:54:12 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.4/8.8.4) id CAA12906 for hackers-outgoing; Fri, 3 Jan 1997 02:54:12 -0800 (PST) Received: from irz301.inf.tu-dresden.de (irz301.inf.tu-dresden.de [141.76.1.11]) by freefall.freebsd.org (8.8.4/8.8.4) with SMTP id CAA12895 for ; Fri, 3 Jan 1997 02:54:08 -0800 (PST) Received: from sax.sax.de (sax.sax.de [193.175.26.33]) by irz301.inf.tu-dresden.de (8.6.12/8.6.12-s1) with ESMTP id LAA27031; Fri, 3 Jan 1997 11:51:27 +0100 Received: (from uucp@localhost) by sax.sax.de (8.6.12/8.6.12-s1) with UUCP id LAA04581; Fri, 3 Jan 1997 11:51:26 +0100 Received: (from j@localhost) by uriah.heep.sax.de (8.8.4/8.6.9) id KAA13966; Fri, 3 Jan 1997 10:09:52 +0100 (MET) Message-ID: Date: Fri, 3 Jan 1997 10:09:52 +0100 From: j@uriah.heep.sax.de (J Wunsch) To: freebsd-hackers@FreeBSD.ORG Cc: mark@putte.seeware.DIALix.oz.au (Mark Hannon) Subject: Re: ft < 50kb/s ?? References: X-Mailer: Mutt 0.55-PL10 Mime-Version: 1.0 X-Phone: +49-351-2012 669 X-PGP-Fingerprint: DC 47 E6 E4 FF A6 E9 8F 93 21 E0 7D F9 12 D6 4E Reply-To: joerg_wunsch@uriah.heep.sax.de (Joerg Wunsch) In-Reply-To: ; from Mark Hannon on Jan 2, 1997 20:28:59 +0000 Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk As Mark Hannon wrote: > ..., however in > the current configuration it is dead slow - It just took me 12 hours > to back up 200m of data, according to the manufacturers notes it You mean, it got slower over time? Even with a new (and freshly formatted) medium? This would be fairly surprising, since the driver hasn't changed except of a few side-effects from global changes. It's basically orphaned (though maybe, i've found a new maintainer). > should be able to do 1Mbit/second, which is around 30minutes. The 1 Mbit/s is, of course, a marketing gag. Since you're operating on a standard floppy controller, it will only do 500 kbit/s, since that's the highest clock rate of a normal (i.e., not 2.88 MB capable) floppy controller. And, you have to subtract the overhead of floppy sectorization and ECC error correction. The floppy sectorization overhead can be as high as (2 MB - 1.44 MB) / 2 MB. All of this, of course, holds only valid as long as the driver keeps the tape streaming. Our floppy disk driver can maintain a streaming data rate (for sequential transfers) of ~ 30 KB/s. Given the design of the `ft' driver, i would be fairly surprised if it maxes out at a much higher rate at all. Of course, the most important question for you is: does the tape remain streaming? If not, all bets are off, and performance will simply suck. We might then try profiling the driver or the [l]ft utility. However, if it keeps streaming, there's not very much we could do. Support for 1 MBit/s floppy controllers might double the rate, i have this on my whiteboard for a long time (for floppy disks, but i figure that it might be easy to have the floppy tape driver also benefit from this). However, somebody mysteriously always puts a new sheet with other ``more important'' tasks over that sheet on the whiteboard... :) -- cheers, J"org joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ -- NIC: JW11-RIPE Never trust an operating system you don't have sources for. ;-)