Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!uwm.edu!ux1.cso.uiuc.edu!csrd.uiuc.edu!sp1.csrd.uiuc.edu!kai
From: kai@sp1.csrd.uiuc.edu (Kuck And Associates)
Newsgroups: comp.sys.sequent
Subject: Re: info about 8mm drive on Sequent
Message-ID: <1990Jul31.185516.8125@csrd.uiuc.edu>
Date: 31 Jul 90 18:55:16 GMT
References: <1990Jul25.160107.4360@csrd.uiuc.edu>
Sender: usenet@csrd.uiuc.edu (news)
Distribution: comp
Organization: UIUC Center for Supercomputing Research and Development
Lines: 23

kai@sp1.csrd.uiuc.edu (Kuck And Associates) writes:
>I'll report next week how monthly full backups go this coming weekend.  I'll be
>using the plain old "dump" command that came with our Balance to write to the
>8mm drive, now that it's a local device, and we'll find out if Sequent's
>"hdump" program really is neccessary.

The backups (using plain old "dump") went very well.  I used the following
options on the "dump" command:
	dump 0unsdbf 300000 6250 32 $DEVICE $FS
Since I never try to dump more than the capacity of the tape at one time, the
idea here is simply to give dump large enough numbers that it never thinks the
end of tape is reached.

It's always a surprise to see the time difference between running "rdump" and
"dump".  Dumps ran in approx 1/3 the previous "rdump" elapsed time.  Tape drive
throughput worked out to around 10 to 13 Mb per minute, even for the disks
connected to the old "xp" Multibus disk controller.  Dumps from other hosts
took the same amount of time as normal.

	Patrick Wolfe   (pat@kai.com, kailand!pat)
	System Programmer/Operations Manager, Kuck & Associates
	"Any opinions expressed are my own, not my employers's.
	 Please don't call my boss and complain ... again."