In the message dated: Thu, 12 Feb 2009 14:22:28 EST,
The pithy ruminations from Doug Hughes on 
<Re: [lopsa-tech] Tapes, not backup...> were:
=> Tim Kirby wrote:
=> > Thanks for the comments so far. As I noted (or tried to) this is
=> > a second hand request and is specifically *not* backup - it's
=> >
=> >   
=> >> tape library management apps for a linux box with an
=> >> attached library for simple tape archival use.

Hmmm....I second the recommendation of "mtx" for managing the tape library 
hardware, but that's way too dumb to help manage the data. If you're dedicating 
this hardware to the "archive", what's wrong with using a backup system 
(Bacula, Amanda, etc.) and simply setting the backup interval and retention 
times so that you get a single full backup that's kept "forever"?


=> >>     
=> >
=> > We're talking about people who want to put a bunch of data
=> > on a tape that may or may not be needed in 5 years time.
=> > Perhaps 10. The "archival" bit is relevant; it's not a "backup".

Explain the difference in this context, please.

=> >   
=> ...
=> 
=> Have you considered 2 tapes? Dropping a tape and having it damaged (or 
=> somebody doing it accidentally at whatever archive site is under 
=> consideration) can be the kiss of death to an LTO tape (for instance). 
=> If you damage the edge of the tape, it can become unusable.

Absolutely. Good advice.

I've been in several situations over the years where someone hands me a box of
tapes[1] and wants to restore data from the "archive". The most common problem
I've seen is not that the media is unreadable due to damage (though that's
happened), but that the logical arrangement of data is unknown or that
hardware that can read the archive isn't available.

With the pace of change in computer systems, 5 years is a long time to be able 
to read an arbitrary piece of media. One advantage to integrating this 
"archival" storage into an existing backup mechanism (you _do_ have backups, 
right?), is that people tend to migrate data from one backup technology to 
another, whereas archives get forgotten until someone wants the data.

As far as the archive...be explicit about documenting the format...for
example... was the tape written with tar or cpio or dd or zip or something else?
Was it GNU tar (which strips the leading "/" and preserves path names over 256
characters) or a legacy vendor tar program? What was the blocking factor (if
any)? Was there compression used? Also, it's very helpful to document the
content of the archive (a full directory listing if possible, a list of the
directory names if the file count is too large, or at least a textual
description of the data). Did the data come from a system that preserves case
differences in file names? Etc., etc., etc.


Mark


[1] Starting with 9-track, 1600bpi tapes, then some random floppies, a few
years later it was QIC tapes and Zip cartridges, then lots and lots and lots of
DDS1, DDS2 tapes, moving on to some CDs, and most recently--just a few weeks
ago--a DVD archive that couldn't be read due to the logical format of the data.

_______________________________________________
Tech mailing list
[email protected]
http://lopsa.org/cgi-bin/mailman/listinfo/tech
This list provided by the League of Professional System Administrators
 http://lopsa.org/

Reply via email to