From owner-freebsd-hackers Thu Aug 29 07:11:34 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id HAA09764 for hackers-outgoing; Thu, 29 Aug 1996 07:11:34 -0700 (PDT) Received: from cyclone.degnet.baynet.de (root@cyclone.degnet.baynet.de [194.95.214.129]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id HAA09738; Thu, 29 Aug 1996 07:11:18 -0700 (PDT) Received: from neuron.bsd.uni-passau.de (ppp2 [194.95.214.132]) by cyclone.degnet.baynet.de (8.6.12/8.6.9) with SMTP id QAA00441; Thu, 29 Aug 1996 16:22:54 +0200 Message-ID: <3225C003.2481@degnet.baynet.de> Date: Thu, 29 Aug 1996 16:06:27 +0000 From: Darius Moos Reply-To: moos@degnet.baynet.de X-Mailer: Mozilla 3.0b6Gold (Win95; I) MIME-Version: 1.0 To: "Jonathan M. Bresler" CC: freebsd-hackers , FreeBSD-questions Subject: Re: [Q]: formula for calculating BPI needed References: <199608290016.RAA15559@freefall.freebsd.org> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-hackers@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk Hi, (previous discussion appended) sorry, but i do not understand your calculation ... because: for B = 500000 and b = 40 KB the number of bytes i would expect of this is 1024 * 40 * 500000 = 2.048 * 10^10 ~ 19.1 GB am i wrong with my calc ? The reason i had the idea of rewriting the formula (and only the formula) was, that the parameters B and b must always be adjusted depending on the type of DAT-tape that will be used. Furthermore, the user himself has to adjust them and calculate accurate numbers. On the other hand the linear BPI for DDS-2 tapes is always 61000 and the length of the tape is printed on the package of the tapes. So using the linear BPI and the tape-length would make users life much easier; but maybe there are other reasons speaking against this method that i'm not aware of. Anyway i would be glad for any response on this topic. It would be nice if a kind soul could point me to sources where the layout of the DAT-tapes and the recording technology is explained in more detail. Thanks in advance. Darius Moos. email: moos@degnet.baynet.de Jonathan M. Bresler wrote: > > Darius Moos wrote: > > > > My first basic question is, how one would calculate the BPI > > for a given DAT-tape-length and capacity. My problem is: > > - let a DAT-tape be 295 feet long <=> lf = 295 > > => length of this DAT-tape is li = lf * 12 = 3540 inches > > - let the capacity of this tape be 2 GB > > => BPI = 2 GB / li = 2 GB / 3540 inches = 606633.8 BPI > > The maual i've read says the BPI is 61000 for such a tape. > > So what am i doing wrong ? > > trying to use bpi for a DAT tape ;) > bpi is from the days of 9-track tapes. use B and b instead. > > B -- number of dump records > b -- number of kilobytes per dump record. > > for a 2GB tape try 500000 40 > > dump Bbf 500000 40 /dev/rst0 > > > > > i was trying to dump filesystems to a SONY-SDT-7000 with dump. > > My problem is, that dumps estimated tapes for the dump are wrong; > > for example dumping /dev/sd0a (about 50 MB) to a 295 feet long > > DAT-tape makes dump to estimate 0.4 tapes (i set the commandlineoptions > > to "0ufds /dev/sd0a 61000 295"). This can not be correct, > > the trick there is helical scan. the bpi may acutally be > 61000. that 61000 bpi are recorded at an angle of ~6 > degrees from horizontal. the width of hte recorded path is > very narrow. many recording paths lie stacked above one another. > similar to repeated slash symbols at a shallower angle "////" > as a result the effective length of a tape is increased many-fold. > > > since the capacity of a 295 feet-DAT-tape is about 2 GB. > > After scanning the sources of dump for the estmated-tapes-formula, > > i found some strange constants being used there that i do not > > understand and that are not documented nor explained anywhere. > > I want to rewrite the formula for this calculation in > > the dump-sources but i need some informations on this topic. > > no, you dont ;) well maybe you do but others dont want you to :) > the formulas in dump are there for the old 9-track tapes. > just use the newer parameters: B and b. > > jmb > -- > Jonathan M. Bresler FreeBSD Postmaster jmb@FreeBSD.ORG > FreeBSD--4.4BSD Unix for PC clones, source included. http://www.freebsd.org/ > PGP 2.6.2 Fingerprint: 31 57 41 56 06 C1 40 13 C5 1C E3 E5 DC 62 0E FB