Path: utzoo!utgpu!watserv1!watmath!att!att!dptg!ulysses!andante!mit-eddie!snorkelwacker!usc!sdd.hp.com!samsung!munnari.oz.au!uhccux!uhunix1.uhcc.Hawaii.Edu!pilger
From: pilger@uhunix1.uhcc.Hawaii.Edu (Eric Pilger)
Newsgroups: comp.protocols.nfs
Subject: Re: lifeline-backup
Message-ID: <9929@uhccux.uhcc.Hawaii.Edu>
Date: 18 Oct 90 19:58:37 GMT
References: <1990Oct16.175419.10751@murdoch.acc.Virginia.EDU>
Sender: news@uhccux.uhcc.Hawaii.Edu
Organization: University of Hawaii
Lines: 31

In article <1990Oct16.175419.10751@murdoch.acc.Virginia.EDU> mlj8e@mljsg.pharm.Virginia.EDU (Michael Johnson) writes:
>
>   I have a problem with Lifeline backups with PC-NFS (sun) and
.
.

>4D80 server.  Everything works fine if I write to a cartridge tape,
>but fails with an LL011F error when I try to write to an Exabyte
>8mm tape.  However, I can read the Exabyte tapes with no problem.
>
I ran into a problem where things would work for awhile, then fail at
the same point each time.  The error was something like LL011F.
I believe the problem was timing out on the write request.  When I
switched from the failing Exabyte on a Sun 3 to an Exabytre on our Sun
490, things worked.  Things also worked writing to a 1/4" drive on a
Sparcstation.  There seems to be some interplay between the speed of
the machine, the type of tape drive, and time outs. (I am guessing I
could also have fixed the problem by increasing the timeout tperiod
for an NFS request.)

>    Can someone out there tell me the actual sequence of remote
>shell commands with the PC-NFS tar (from SUN) is actually sending
>to the server?  Can anyone offer and suggestions?

They are stored in backup.db in the NFS directory.  They are
something like "rsh dd if(of)=/dev/(your tape) bs=(your block size)."


				   Eric Pilger 
				   Systems Programmer
				   Nasa Infrared Telescope Facility