Backup vs. Archival

With the recent announcement of Apple’s Time Capsule product, and the encouragement from Steve to his users to backup their stuff, I thought I’d take a moment to clarify a distinction between what backup means and why you might (in some cases) be archiving rather than “backing up”.

A backup is an extra copy of something currently on your computer, stored in another place. It’s purpose is so that if one of those locations fails (the source or the backup location), you still have one copy and you haven’t lost anything.

Archiving, on the other hand, is the storage of your data on another device (or burned media, etc.) for the purposes of keeping it around for a while. An example of this would be if you have a bunch of large files for a project you’ve been working on on your laptop, and now that you’re done with that project, you move those files off to your network attached hard drive, to free up the space on your laptop.

I’ve often heard the latter referred to as “backing up”, but of course it isn’t, because your data is still only in one place, and you will lose it if that network drive fails. What you really want is to be sure that for all your data (both your “archived” data and the stuff you use every day) to be backed up.

On a related note, while the Time Capsule is a cool device, I think it would have been good if they included RAID capability. That way if the HD in the capsule dies, you haven’t lost everything. This would especially affect people who are archiving stuff to the network drive and thinking of that as their “backup”, because then their data would be gone.

If you’re looking for network storage for the purposes of backup and/or archival, I recommend a cheap Linux box that you can load up pairs of hard drives into as needed, using RAID 1 or 5 to provide redundancy. This way, if a drive fails, you just throw in a new one and let it resync, and you haven’t lost any data. Fry’s frequently has sales on hard drives; these days you can often find 500GB for under $100. I’d choose Linux’s software RAID support too, rather than going with a RAID card or built-in motherboard RAID support; I’ve seen people get stuck when their decade old RAID card fails and they can’t find a replacement because they don’t make them anymore.

If you do the RAID thing, you’re pretty safe from drive failures, but for data you really care about, you should probably still have a separate backup copy, preferably off-site, to account for things like theft, fire, or other damage that might affect your local storage.

13 Comments

  1. Posted January 19, 2008 at 2:02 pm | Permalink

    Let me correct you so you can update your post and not mislead your readers.

    Time Capsule is not just a NAS but it’s primary function is an add-on to Time Machine. Time Machine will back-up to the NAS all of your files. It not only stores the latest revision but it stores the last X-days revisions as well.

    The great thing about time capsule is the ability to backup notebooks. Prior to the time capsule you had to attach the notebook to a USB or Firewire drive in order for the backup to occur. Now it can be done wirelessly through the home network.

    In the near future Time Machine will allow for full OS backup through Time Machine but if your main drive crashed now it’s simple enough to reinstall the OS and recover from the Time Machine backup.

    So time capsule is not just a NAS to backup or archive to, if you use it as it with it’s primary purpose, with Time Machine, you will be safely backed up.

    I would also say that RAID is overboard if you’re using an external drive or NAS. Because if one goes out you have the other (original or external). I don’t disagree with being more redundant, I’m backing up to an external drive and I have RAID setup through OSX’s awesome RAID support, but for the average user who might not know what to do when the array fails backing up to an external drive is fine.

  2. Posted January 19, 2008 at 2:06 pm | Permalink

    Maybe you’re not saying time capsule isn’t used for backing up. Were you?

  3. Posted January 19, 2008 at 2:53 pm | Permalink

    I think maybe you missed the points that I was making, because in reading your comment I didn’t see any corrections to anything I said. I definitely don’t think I said anything that could be misleading.

    It’s pretty clear that Time Capsule is meant for backing up. I never said it wasn’t.

    The problem I was alluding to was when users use it (or any other networked drive) as a NAS - in other words, not using Time Machine or other backup software, just moving files over to it.

    The point was that if you move files off of your laptop onto a NAS (which many people do because most laptop drives are not super large) you do not have a true backup.

    RAID is definitely not overboard if you’re storing files on the network drives that you want to keep. Again, I’d bet that most people using NAS devices from laptops are using them to store files that they don’t have room to also keep on their laptops. The point I was trying to make is that if they’re doing that, they aren’t backed up, and RAID is a cheap, easy thing that could help them in that situation.

  4. Posted January 19, 2008 at 3:36 pm | Permalink

    It’s a rarity that you start out a post about an apple product and Steve Jobs that you’re not dissing the product, that’s why it seems like your saying that time capsule doesn’t or it’s intended purpose is not to back-up and that Apple is misinforming it’s customers.

    Since your not, it’s good that you’re pointing out that Time Capsule does backup right.

    I still only think RAID is good for archiving on a pro-consumer level. Otherwise having an external RAID has less stability than an external drive backup and a DVD or CD backup. RAID is easier day-to-day but setup, administration and recovery isn’t easy.

  5. Posted January 19, 2008 at 3:37 pm | Permalink

    Reading it again, I guess I could see how it could be taken as a slam on Time Capsule. I didn’t intend it that way, just referencing it as something that recently got a lot of people thinking about backing up in general. So, backing up was the main point of the post, not the Time Capsule.

    Other than the comment about how I think it should have offered RAID (which I still do), I think the original post was generally pretty positive about Time Capsule.

  6. Posted January 19, 2008 at 3:57 pm | Permalink

    I wish comment date stamps were on page load and not submit, that last comment should be before your last. (Not that it took me that long to write but I just came back to the page after doing something)

  7. Posted January 19, 2008 at 3:59 pm | Permalink

    I’ll adjust it.

  8. Posted January 19, 2008 at 4:05 pm | Permalink

    I disagree on the RAID point. It’s very easy to configure, cheap, and has huge benefits over having to always burn everything to DVDs.

    Recovery is easier with the RAID model too; all you do is throw a new disk in, for Linux at least. Maybe Mac OS makes it hard for some reason. :)

    Also, there’s nothing stopping you from still doing DVD or other more “permanent” copies of the data too, as an additional failsafe, but the availability advantages of RAID are hard to argue with.

  9. Posted January 20, 2008 at 9:19 am | Permalink

    RAID is not easy to setup the first time on Linux because there’s too many options and ways to screw it up.

    And no, MAC OSX RAID is as easy as it gets because it has simple defaults. The last time I worked on a Linux RAID (1 year ago) it had no defaults so learning about how it worked was a chore.

  10. Posted January 20, 2008 at 12:53 pm | Permalink

    I guess it’s a matter of opinion; I think it’s easy enough to be accessible to anyone who wants to set it up.

    After thinking about Time Machine a bit more, I do have a question that I wonder if you can answer, specifically regarding the version history feature:

    When you configure Time Machine to use Time Capsule to store the data, is that a secondary storage location for that data, and is it all still available on the laptop as well (without the Time Capsule)?

    If not, I would say that (without also backing up the Time Capsule data to other media, if you can even do that) it’s definitely incomplete as a backup solution.

    I also wonder whether Time Machine on its own (without Time Capsule) supports storing an additional copy of its data somewhere other than (and in addition to) the computer it’s running on. If not, it’s also incomplete as a backup solution.

  11. dan
    Posted January 20, 2008 at 3:49 pm | Permalink

    I disagree with it being a matter of opinion, maybe experience or commitment. Either way I wouldn’t tell anyone in my family or every one of my friends except for a few to use raid unless its one of those dedicated devices that handle it for you.

    I’d have to confirm if time machine has the version history locally, I would say maybe not since that’s a lot of data.

    I really don’t get your point though because while your picking apart TM your exceeding your expectations for backing up. Because you no matter you have a backup of your current files. And if you want to backup your time machine backups you’re allowed to.

    No matter how you feel about apple you shouldn’t discredit TM, its something every OS should strive to provide to its users by DEFAULT. And it transcends backing up files to an external drive.

  12. dan
    Posted January 20, 2008 at 3:54 pm | Permalink

    Two things:

    TM does work locally, what I need to confirm is if you backup to an external drive if it continues to keep the copy locally and on the external drive. My guess is it does, since TM always works on a laptop even if you use TC to backup.

    I’m on my iPhone.

  13. Posted January 20, 2008 at 4:19 pm | Permalink

    RAID:
    All I can tell you is that I’ve set up RAID configs several times, and (at least in some of the latest distros) it was incredibly easy. Apparently your experience was different, but I wouldn’t use that as a basis for saying it’s categorically hard across the board.

    Time Machine / Time Capsule:
    I’m not picking it apart or discrediting it, just talking about it in the context of my original post. You seemed to be more interested in how it related to Time Machine / Time Capsule specifically, and so I was continuing the conversation along that line.

    My only point as it relates to this post was that if there’s data (such as the version history) that TM only stores in one place (whether that’s locally on the laptop or on the TC), then that data isn’t backed up unless you manually take other steps to do so.

    The point I was trying to make in the original post is that in any backup solution (TM or otherwise), if you have a single point of failure, you don’t have a true backup.

    I’m not saying TM/TC is a piece of crap or anything. I generally think it’s a great idea, but I’m just suggesting that to truly make it a good, reliable backup solution there are some improvements that could be made.

Post a Comment

Your email is never shared. Required fields are marked *

*
*