Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: REPOST: RMAN question

Re: REPOST: RMAN question

From: Howard J. Rogers <hjr_at_dizwell.com>
Date: Tue, 31 Aug 2004 15:44:24 +1000
Message-ID: <41341051$0$22903$afc38c87@news.optusnet.com.au>


Daniel Morgan wrote:

> Howard J. Rogers wrote:
> 

>> Daniel Morgan wrote:
>>
>>
>>>Howard J. Rogers wrote:
>>>
>>>
>>>>And they're a lot more safe on tape "forever" than they would be on
>>>>disk.
>>>>
>>>>HJR
>>>
>>>Not my experience. On what is your statement based?
>>>
>>>Thanks.

>>
>>
>> On having formatted more hard drives than I can shake a stick at, and
>> also on having happily zip-locked plenty of tape cartridges and sent them
>> off to a bank vault.
>>
>> Regards
>> HJR
> 
> I'll agree with all of your comments and disagree with your conclusion.
> 
> I too have formatted a lot of hard disks. But never one with static
> content put away for safe storage.

I can remember us putting a hard disk away once. Not for archiving purposes, but with the intent of sometime sending it away for a bit of data recovery. I caught a junior fetching it out from the filing cabinet one evening, intending to install it into a user's workstation since their own hard disk had failed...

> Now the question that really matters. What makes you confident that
> those tapes you sent off were usable for recovery 5+ years later? 10+
> year later? Or more?

I doubt the tape drive we recorded them on would still be usable today. That's why we transferred stuff which really did have to be kept "forever" (and not just for many months) onto CD (as it then was) and DVD (as I would probably do today).

This isn't rocket science. People have had to discuss the issue of long-term storage of critical data many times over the years (and let's leave aside the fact that I doubt a program exists these days that could read the data I considered critical back in 1987). If you want storage that will last for millenia, get out your chisel and hammer. Short of that...

Of all the possibilities for long-term (ie, many, many months) storage, I don't think I've ever heard a hard disk seriously suggested. That could just be me, of course. Or it could have something to do with the fact that there is no write-protect tab for a hard disk (see note above). That one drop on the server room floor toasts the disk. That an airline of my acquaintance is very iffy about transporting hard disks it can't plug in (and potentially destroy in the process), but is not so concerned about a DVD or CD-ROM. And so on. Hard disks are big and bulky. CDs are not. Tapes are fairly robust; hard disks are not.

I can think of no medium which screams out "I am short term" more than a hard disk. Nor one which better cries "I am a reasonable medium-to-long term proposition" than a tape.

And I know you won't believe me, so try believing an institution such as the British Museum whose very existence is predicated, to some extent, on making these sorts of decision. To which end you might read

http://www.ukoln.ac.uk/services/papers/bl/jisc-npo50/bennet.html

And in particular, section 5. The opening sentence of which reads

"In the last 50 years, the diversity of media on which data has been stored has not diminished, but increased. Despite the diversity, the most durable of media remains the tape."

As they go on to conclude: "...it is the opinion of the team that there is no real choice over the ideal media for long term preservation of digital material. The media that should be used is either 8mm DAT volumes or some derivative of CD ( a new CD format may require bulk copying of data)."

Despite their lack of hard conclusions, you might note that the hard disk did not feature in their final choices.

Now you might conclude that the paper dates from 1997 and is therefore worthless (and I'd agree with you regarding their conclusion that Windows 95 and SCO Unix were safe long-term bets!). But I could post you other papers from similar institutions of more recent vintage that come to much the same conclusions.

Regards
HJR Received on Tue Aug 31 2004 - 00:44:24 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US