Long-Term Personal Data Storage?

BeanBagKing writes "Yesterday I set out in search of a way to store my documents, videos, and pictures for a long time without worrying about them. This is stuff that I may not care about for years, I don't care where it is, or if it's immediately available, so long as when I do decide to get it, it's there. What did I come up with? Nothing. Hard Drives can fail or degrade. CD's and DVD's I've read have the same problem over long periods of time. I'd rather not pay yearly rent on a server or backup/storage solution. I could start my own server, but that goes back to the issue of hard drives failing, not to mention cost. Tape backups aren't common for personal backups, making far-future retrieval possibly difficult, not to mention the low storage capacity of tape drives. I've thought about buying a bunch of 4GB thumb drives; I've had some of those for years and even sent a few through washers and driers and had the data survive. Do you have any suggestions? My requirements are simple: It must be stable, lasting for decades if possible, and must be as inexpensive as possible. I'm not looking to start my own national archive; I have less than 500GBs and only save things important to me."

Comments

Not enough history

[mgkimsal2]

We don't have enough history on this tech to know what, if anything, will "last for decades". Possibly "paper" and "microfiche" might fit in that list, but those aren't the sort of things you're talking about. Best option I can think of right now would be to get a couple 500gig drives, put everything on both, and then put them in different areas. In 3-5 years, back them up to something newer, and repeat that every 3-5 years. Maybe in those intervening years, we'll have more data and newer tech that's demonstrably suited for what your needs are.

Re:Not enough history

[gad_zuki!]

One drive should be "live" and the other archived. Considering we all own computers, throwing a 1tb drive into a box isnt so difficult. Hell, you could write a script to power it up once a month and then power it down, if people are worried about energy costs but dont want to keep it spinning 24/7. It doesnt need to be ever mounted.

Better yet both disks should be running in a RAID 1 array. This is a cheap solution, but its not a "toss in the closet and forget" solution. If this guy actually cares about his data I dont see why he cant spend 200 dollars or so for two drives and a raid 1 card.

I see this question at slashdot every couple of months. The answers are still the same. Keep it live on a disk until a better solution is found. Upgrade the disk every so often. That's it. Mods, stop posting the same damn question every month.

Not being answered

[suso]

Obviously this question hasn't been answered for the general public because this is like the 4th year in a row that this question has been asked on Slashdot.

Re:Not being answered

[sjames]

That's because there is no fully satisfactory answer. We'd all like a just do this, throw it in the corner and when you come back for it in 50 years it'll all be there sort of solution, but there is no such beast within the realm of affordability.

It's a problem with several aspects to it as well. Let's say there is a SATA drive out there that absolutely CAN sit in a safe deposit box for 50 years and then work perfectly every time. In 50 years, all computers will have whatever the successor to whatever replaces SaS and when you mention SATA, the old timers will all get nostalgic and go on about tying onions to their belts (which was the fashion at the time). You'll then have to take the decidedly NOT affordable step of having someone build you a one-off SATA controller that can interface with a computer of that time. That is, if you can get the old-timers to stop reminiscing about the Vista debacle of aught eight long enough to recall the specifications of SATA. Be sure to duck, some of them might throw a chair for ilustration.

Not the media that's the problem

[jwilkins13]

Here's the thing. Flash drives will *probably* last long enough. I wouldn't at all be surprised if they were still readable in 20 or 30 years. But a) what's the odds of your current WinUx or iHoloTablet having a usb connector in 30 years? and MUCH more importantly, what's the odds of having anything capable of reading those historic Word 2007, Acrobat 5, or any other type of file format in 20 years? Yes, there are some folks technical enough that they can still read and readily interact with Geoworks, Wordstar, Xywrite, etc. stored on 8" floppy disks. But if you ain't one of them, and I'm happy to admit I ain't, the fact that the flash drive is physically capable of being read in 30 years simply won't matter. That's why I crack up reading various vendors' claims of CDs, DVDs, BDs etc. lasting 50 or 100 or more years. The disks will be readable but you'll have no mechanical or logical way to read what's on them.

An archive is not a long-term backup

[ewilts]

There is absolutely nothing that you can put away for decades and expect to be useful. Your requirements are not simple - they'll actually very, very hard to meet, even if you want to throw a lot of money at the problem.

You don't know that a jpeg, for example, will be readable in 30 years. The format may be so deprecated that there might not even be a viewer available. Like my old Microsoft Works 4.0 documents - although I have the data, I have nothing that can read them unless I want to spin up an old Windows image, assuming that I can generate a virtualized environment that can support an old Windows (Windows XP probably won't even boot on any PC being produced 30 years from now). And some of that data is only a few years old, not decades old.

You should store not only the data, but also the applications that created the data. And the computer you need to run those applications. And backups of those. And then every few years, pull it all back and validate it and update as required.

You may have only 500GB now, but 10 years from now that will be 5TB. And then you need a way to actually be able to find something you added to your "archive".

I deal with this at work regularly. An archive is not a backup that you keep for a long time. It's much, much more than that. Once you start thinking about all of the issues that come up, you'll see that the media is the least of your problems.

Re:An archive is not a long-term backup

[eean]

I think we do know that a JPEG will be readable in 30 years. Formats that have been around for like 10-20 years like JPEG are going to be here for a long while longer; I'd say until the end of civilization at a minimum (and even then, it wouldn't be hard for people to figure out the format). The worse case is that in future generations only a librarian or data archaeologist would have the tool to open it. Given the open source nature of JPEG, more likely you'll just download a JPEG viewer.

MS Work 4.0 documents is completely different. There was always only one implementation, it wasn't open source, it wasn't a documented standard, and the life span of the format was small to tiny.

Re:An archive is not a long-term backup

[nine-times]

I think you're exaggerating the problem a bit. Formats like GIF, JPEG, and ODF will most likely be readable somehow in 30 years. They may not be the format of choice, but we have open source readers for those things, so for as long as lots of people have data in those formats, someone will be maintaining viewers that allow reading them and probably converting them to newer formats. Besides, it's not clear to me that we're going to come up with much better compression methods for static images, or that we really need to bother coming up with much better compression methods for static images, which means it isn't that unlikely we'll still be using JPEG in 30 years. I'm not saying it's a lock or anything, but it's not *that* unlikely.

Now, with a format like ODF, if adoption isn't bigger before something new comes along, you might have a hard time reading that just because of the relative obscurity of the format (which is a problem JPEG doesn't have). In that case, it will probably depend entirely whether enough people have enough valuable information in ODF that some developers somewhere think it's worth writing a viewer.

Yes, ideally emulation would be available for every obsolete platform, and we'd all keep VM images of all our old operating systems. We'd all keep all of our old applications to install on those images, and VM software would always be backwards-compatible meaning that we'd never lose anything. I'd love to know that someone somewhere is working on that, if only for historical preservation. However, for the individual who might have limited resources, it probably won't be necessary. If it ever becomes necessary for that to happen for most people, someone will be able to make a lot of money selling a solution.

In most cases I'd say the best bet is to stick to open formats, keep copies on multiple different media, and continually migrate to new media. So, for example, back everything up to a hard drive and create checksums for every file, and then burn multiple copies to DVD. In 3 years, pull them all out, check all the checksums for corruption, and copy known-good copies (and checksums)to your brand new 5TB hard drive, and burn a couple BluRay discs. In another 5 years, check the checksums again, get known-good copies, and copy them to your 50 TB SSD and burn a couple copies into your super-ultra-cool whatchamakallit.

It doesn't exist.

[NReitzel]

You are the analogy of an investor who wants a high-yield, low-risk, completely liquid instrument. The term is TANSTAAFL.

I maintain two (yes, two) USB external drives. Every couple of years, I migrate to a larger, or otherwise better medium. I use an incremental backup system (for me, cpio) that ends up keeping too much stuff, but at least I have the stuff I want if I need to get to it.

In a decade - in my case, four decades - one can accumulate a remarkable amount of crap, along with things one truly wants to save. I have a total of about 90 gig of actual data, plus a far larger amount of music and video, which I consider more or less disposable. It is not difficult, nor expensive, to purchase another external drive and copy the data. My oldest backup is on IBM 2314 disk pack, but the data still held on that disk is also present on my current backup, a WD 160G in a USB-1 enclosure. Sometime next year, I'll go to a 500 G drive in a USB-2 enclosure.

An important consideration is to periodically check to see that the data ostensibly held on a drive (or CD, or DVD) is actually readable. DVD/RW in particular has a tendency to get flakey over long periods of time, expecially if stored under adverse conditions (jammed in back of desk drawer, under sixteen pair of scissors, stapler, a box of pop-tarts, and four old coffee cups. I always keep my last few generations of backups, and if I find an unreadable datum, I make an effort to recover it from the previous backup.

While it may be stating the obvious, it's a Bad Idea (TM) to wait to back up data until you have a problem. I back up all of my data every week or two, and critical data, daily, without fail. Critical data is cached as a three-generation dataset (IBMese).

Good luck. There are no real solutions, just ways to cope.

Changing formats

[DynaSoar]

[From Slashwayback]: Dear Keypunch, I have data I want to keep for decades. Should I invest in a good card reader, or should I transfer my data to these far more efficient but newfangled "floppy disks"?

It's pretty ridiculous to expect one storage format to be viable for 'decades'. Not because it goes bad (even though it probably does), but because you're not likely to be able to maintain the necessary equipment for that long. If you find a storage solution, you need a retrieval solution to equal it. What equipment will you be able to find decades on that can access your storage, even if it stays good? You have no idea.

I've been maintaining a collection of Apple IIs and recopying the programs and data regularly (mostly through full HD backup, reformat with error block deletion, reformatting and replacing) to keep it readable. I have machines and data between 20 and 30 years old. I recognized long ago this had become a hobby in its own right, as most of what I had hasn't been of interest to me for many years. The little bits that have been useful have been transferred to newer machines and formats several times. That's decreased as more and more of it can be found easily on the web (previously FTP/gopher/etc.).

Get used to transferring your data to new formats as they come into widespread use, and recopying as necessary to keep them readable. Or else:

[From Slashwayforward] Dear Galactic EM Field Computing, I just found about 20 pounds of aluminized plastic disks that used to have data on them, but I can't read them to tell if I still want it. Is there any museum that might want these? Or are there still any operating plastic recycling centers that might give me a few bucks for them?

Think Different

[crdotson]

I think the issue is that people are thinking about this incorrectly. You don't really want to 'archive' this data -- keep it with you! Keep it with all of the data that you are using day to day and back it up and move it along with that.

My home workstation still has files from 15 years ago on it. I've replaced the computer many times, had a few hard drives fail, etc. but I've always restored both current and 'archive' data from backups and kept going.

Re:Amazon S3

[bcrowell]

Well it's not dirt cheap for 500GB - that's $75/month.

The OP specified "decades if possible." So we have two problems here. One is that 30 years at $75/mo comes out to $27,000, which is a little pricey. For that amount of money he could probably hire someone to come to his house once a year, verify the readability of his media for him, and transfer them to new media as the old ones become obsolete. $1,000 for a few hours' work? I'd take it.

The other problem is that the probability that Amazon S3 will exist in 30 years is very low. This is basically the problem with any possible answer to his question. There isn't any computer-related service or equipment that you can be sure will still be there in 30 years. A more realistic goal would be to do it in 10-year steps; if that's all he wants, then the shoebox full of flash drives should work fine, and then 10 years from now he can transfer those data to something else.

But who really has 500GB of critical data.

I initially misread that as 500 Mb, which is about the amount of critical data I have that needs backing up. 500 gigabytes is kind of a crazy amount of data. One way to get that much might be that he has a gigantic collection of mp3s, or possibly a moderately huge collection of music in a less lossy format. But then that's not critical personal data, it's just a music collection. And the chances are that as the decades go by, he'll realize that the music he thought was so important and wonderful in 2008 no longer seems so important to him. I know plenty of people who still have their Kool and the Gang LP's from the 1970's, but it's not like they're willing to spend a thousand dollars a year to obsessively maintain them.

Re:Magnetic Tapes...

[nine-times]

I think your post is very insightful, and I have an additional problem to throw into the mix: sorting through all the crap you've archived, even assuming you can read it all.

I don't know about you, but I've run lots of different backups on lots of different systems, and one of the problems that always comes up is just finding the revision of the file you want. People say, "I want the copy before I made this revision-- I think I did that about a month ago." Check the backups and there are no revisions from a month ago, but there are 20 from the month before. Next thing you know you're checking 20 copies by hand, and none of them are what you're looking for-- and that's even when your backup/archive system is working.

So when devising any kind of archive, I think it's at least worth considering, "How am I going to find what I'm looking for in 20 years?" Imagine yourself in 20 years, and you have every piece of data you've ever generated stored on some kind of media that holds hundreds of terabytes of data. You want to find some spreadsheet you made today (20 years ago). Maybe you don't remember exactly when you made the document-- you think about 15 years ago, but it's actually 20. You can't really remember what the filename was. You can't remember if you made it in Excel or OpenOffice, so you're not even sure what filetype you're looking for. What's going to be your method for finding that file?

I'm not suggesting it's an insoluble problem. It might be that it's not even a problem in 20 years because indexing/searching has become so good that your AI will be able to sort through terabytes in a couple seconds and make some good guesses about what you're looking for, but do you really want to rely on that happening?

Re:Not Amazon S3

[anagama]

Encryption only works because brute forcing the scheme on current hardware is ridiculously time consuming. Encrypting with today's standards does not protect against future advances in computing power.

Re:Not Amazon S3

[MikeBabcock]

If someone's willing to brute-force a password to change your data, they're willing to copy your DVD's ISO, edit it, and burn a new copy to replace it with.

Re:Amazon S3

[vux984]

A lot of people shoot home movies and then become obsessed with preserving the footage. That adds up fast.
It seems that deciding what isn't important is a hard part of backing up.

After a loved one dies, even the lowest quality outtake with a thumb covering half the shot can be a priceless memory.