🚀 go-pugleaf

In article <1992Oct8.170221.14876@riacs.edu> lamaster@pioneer.arc.nasa.gov (Hugh LaMaster) writes:
>Archival and/or Backup - That is, taking a file or a directory tree, 
>and storing it offline somewhere, but where it is accessible.  Two
>key features can make a backup product "archival":  The ability of
>users to archive and restore files without needing root, and the
>ability of users (and root) to see what has been archived.  Otherwise,
>archival is similar to backup.

Apparently the terms "archival" and "backup" are not universally defined,
then, because that's not how we would differentiate between them at all.

IBM is now selling at least one backup/archival product, the Workstation
Data Save Facility, and has announced its follow-on, the Data Facility
Distributed Storage Manager.  Both WDSF and DFDSM provide backup and
archive services for a wide variety of IBM and non-IBM clients, such as
DOS, OS/2, AIX, SunOS, MacOS, NetWare 386, Windows and others.  The same
backup/archive server can support various clients at once, via a variety
of communications protocols (TCP/IP, NetBios, IBM 3270 emulation, etc).

In these products, two separate subsystems are provided.  The "Backup"
system is intended to protect users from both catastrophic system failures
(head crash) as well as user errors (oops, erased the wrong file).  
Users can restore their own files without being root; root can restore
everything.  Users can see everything that has been backed up ... including
multiple versions of each file, files that are backed up to DASD and files
that are backed up to tape, etc.  (Note that we call this the Backup subsystem,
while Hugh would characterize this as "Archive").

The Backup subsystem backs up user files either on a schedule or on demand.
As such it's kind of "active"; it actively stores away versions of the users'
files in case they are needed.  It actively discards old copies of files 
when it deems them unnecessary (if a file is erased by the user, then 
"k" days later the backed up copies are erased, too).

The "Archive" subsystem, by contrast, is a passive, user-driven thing.  
While it's implemented using some of the same infrastructure as backup,
it's quite different when viewed by the user.  Files are archived by the
archive subsystem only when a user explicitly tells the system to do so.
If a user says, "archive file /u/drake/important.stuff for 90 days", then
that's exactly what the archive subsystem will do ... make a copy of the 
file and keep it for 90 days.  No more, no less.  

You can think of "backup" as an invisible safety net for users; if data
is lost or corrupted, they can get it back.  They took no action to get
it saved, and can take action to retrieve it.

You can think of "archive" as the world's smartest tape drive.  It a user
has data that they want to make sure they can get to later ... even years
later ... the user can explicitly tell the system to make a copy and keep
it for "n" days.  Without an archive system, the user might stick a tape
in a tape drive, write the file onto the tape, then throw the tape in a desk
drawer (and lose it).  But the user calls the shots, and the user is always
right.

I'm not saying that these are the "right" definitions for the terms "backup"
and "archive" ... but they make sense to me, and yet are quite different
than what Hugh meant by the terms.  SO, perhaps a useful thread is:

	What do YOU mean by the terms BACKUP and ARCHIVE?


Sam Drake / IBM Almaden Research Center 
Internet:  drake@almaden.ibm.com    BITNET:  DRAKE at ALMADEN