Date: Thu, 1 May 2003 17:03:18 -0500 From: Dan Nelson <dnelson@allantgroup.com> To: "Jacques A. Vidrine" <nectar@FreeBSD.org>, freebsd-questions@FreeBSD.org Subject: Re: huge file, spanning multiple tapes Message-ID: <20030501220317.GA58262@dan.emsphone.com> In-Reply-To: <20030501213842.GA58002@madman.celabo.org> References: <20030501213842.GA58002@madman.celabo.org>
next in thread | previous in thread | raw e-mail | index | archive | help
In the last episode (May 01), Jacques A. Vidrine said: > I thought this would be an easy problem to solve, but I haven't found > a decent solution after a bit of looking. > > I have a huge file, say 73GB. I need to dump it to tapes that have a > capacity of about 33GB each (but there is hardware compression, so > there is no exact capacity). > > If the file would fit on a single tape, I'd just use good 'ole dd(1). > > I figured, heck, somebody is _bound_ to have written a simple utility > that, say, reads from a pipe and writes to tape (or vice versa), and > prompts for a tape switch when it hits end-of-tape. But alas, I > cannot find such a beast. > > Using dump, cpio, etc is not really an option. I need to be able to > later read the data from tape into a pipe without it hitting disk. > > Anybody have bright ideas? How about tar? That lets you specify a change-volume script so you can change tapes when you're writing, and when you read the tape later you can specify the --to-stdout flag so it doesn't write to disk. If you can't have any headers, you can always tell dd to only write 33GB per tape: gzip < file | ( dd of=/dev/ersa0 bs=64k count=$((33000000/64)) echo "please insert next tape" head -1 < /dev/tty > /dev/null ) -- Dan Nelson dnelson@allantgroup.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030501220317.GA58262>