How to back up your dissertation – really.

May 4, 2009 at 6:17 pm 6 comments

Last week one of my friends posted on facebook that they had accidentally erased one of their dissertation chapters. I felt clammy hands, a cold sweat, and a minor chill pass over me. I even think my Krebs Cycle was momentarily disrupted. I have a confession: I live in mortal fear of losing data.

If you look at it just from the economic perspective, the dissertation is probably the most expensive thing most people have written to date. In the shower today I did a mental run down of the total cost of just the dissertation (not the whole Ph.D. Degree). So, at Harvard you are guaranteed a finishing grant that comes with a stipend of $22,330, plus a waiver of the tuition and fees ($4,868). You have to have two chapters written before you can get one.

So let’s imagine the worst case scenario: your dissertation mysteriously vaporizes the day before you turn it in. What is the replacement cost? Well, assuming you wanted to maintain your standard of living, you would have to pony up $27,198 just to cover the finishing grant. But assuming that you wrote at a steady rate, let’s say that the stipend really only covered 60% of the time you needed to write (since in theory you had two chapter (of, let’s say, five) in hand before you started). That’s another $14,886, meaning that the replacement cost of a Harvard dissertation is somewhere north of $42,084. Fair market value is, of course, another question …

In other words: most people can’t afford not to have a data recovery plan. I call it a “data recovery plan” instead of simply backup, because my friend got in trouble precisely because she was backing up. My friend explains:

I realized I should probably give docs a new file name every day – is that what you do? Currently I save on the computer I’m using, immediately back up to a memory stick, and then copy on to either my laptop or desktop when I next get to it, which would be a fine old system if I hadn’t (I assume) one day ‘drag and dropped’ the wrong way. I figured the good thing with this is that my pc and desktop are never in the same place, so I’ve at least got the geographic distance thing taken care of!

For my money, you want your data recovery plan to be as automatic as possible – you don’t want to have to be responsible for it. So mailing CDs or printouts to Mom & Dad (or to yourself) works for a while, but it is awful dreary, isn’t it, to have to do that all the time?

Here are some guidelines that strike me as important, and then discuss my own solution. Some people may already have these in place (although, if you do, shouldn’t you stop wasting time reading blogs and go write?)

1. Automaticity. Ideally, you should never have to take conscious action to perform your backup. If you do, you are introducing human error into the equation, which can be bad.

2. What are you protecting from? There are any number of things that can destroy your data: hardware failure, user mistakes (like overwriting the file), or fire/flood/natural disaster. The most comprehensive data recovery plans try to cover as many concievable situations as possible. Hardware failure happens much more frequently than you might think – if a disk is only warrantied for 3 years, well, that tells you something about what the manufacturer thinks about how long it will last, doesn’t it?

3. What is your tolerance for downtime? If you have a catastrophic loss of data, how quickly do you want to be back up? Will you have to wait for Mom & Dad to send you the printed out copy, and then type it all back in? What about field notes/data/etc? Do you have enough money to buy another computer? Do you have one you could borrow from someone? Do you actually know how to get data out of the media that you are backing up to?

4. How incremental do you want to be? Do you change your mind after writing for a week and then go back to a previous version? Do you just need a snapshot of your most recent stuff?

5. Where do you store canonical data? This might get a “huh?” from people, but is common if you have more than one computer (i.e., laptop, desktop, thumb drive, etc.) My friend got in trouble because the canonical copy of her chapter got clobbered by a non-canonical one.

Here’s how I protect my data. It is a two-tiered solution that was worked so far (knock on wood).  This covers all of my files from graduate school, my media library (I am not going to re-rip every CD that I have in iTunes because a hard drive craps out), dissertation, writing notes, pretty much everything.

1. Automatic Backups via Windows Backup. These are pretty simple to configure, and I splurged and bought a 750GB drive for my desktop so that I could back up everything that I needed. Windows Vista also does all kinds of sexy shadow copies, etc., which I’ve never exactly figured out but can be used to get earlier versions of files …. sometimes. But I also know that a separate disk drive in one computer only protects me from certain kinds of contingencies. Yes, you do need to leave your computer on overnight if you want this to run after the day is done (typically backups don’t do well if you are actually trying to use a computer while they are running – to resource intensive). So it isn’t the greenest solution in the world. But if that bothers you, there are always carbon offsets available for purchase.  A major electrical storm could wipe out a computer and fry the system. And then I would be SOL. Which is why I have also found a particular friend in

2. Jungle Disk (http://www.jungledisk.com). Jungle Disk is a utility that you can use in Mac, Windows, and Unix environments to interact with Amazon’s S3 (http://s3.amazonaws.com). What is S3? It is basically a pay-as-you-go disk drive on the internet. For $.15 per month per gigabyte, you get to hold as much data as you want (really – although I think if you have over 50 TB the costs drop – but that is a very long dissertation).

Jungle Disk is free and easy to configure, although if you pay the $20 you get some nice bells and whistles. I could sing the praises of it forever, but we have a 1,000 word limit here at amusicology and I’m already over. So go get it and save yourself some heartache.

 

Entry filed under: Uncategorized. Tags: .

Harvard Group for New Music; Firebird Ensemble Michael Jackson

6 Comments

  • 1. ravi246  |  May 5, 2009 at 7:27 am

    nice thought…do read me on http://ravi246.wordpress.com/

  • 2. nosila  |  May 11, 2009 at 6:06 pm

    A major electrical storm could wipe out a computer and fry the system. And then I would be SOL. Which is why I have also found a particular friend in

    Ha!

  • 3. Emily  |  May 12, 2009 at 1:43 pm

    Nice, Drew. Very glad if others can learn from my mistake. I’m investigating Jungle Disk as we speak!
    I’d also remind people to be sure to have a backup option that involves geographic distance. I think quite a lot of people think backing up to a thumb drive or external harddrive is enough, but if your computer dies with it in it, or is stolen with it in it, or (heaven forbid) the building burns down (!) you’re screwed. This is where online storage strikes me as an excellent option.
    Thanks!!

  • 4. Emily  |  May 12, 2009 at 2:35 pm

    And PS…
    I don’t think I’ve ever been cited quite so much in one article. Although I was shocked by the alarming lack of footnoted references.
    All the same, it must mean I’m headed for the big time…
    I’m just setting up Jungle Disk – so far, it looks great!
    Thanks again…

  • 5. eliot  |  May 17, 2009 at 10:06 pm

    For mac users, Time Machine is an excellent alternative to Windows Backup.

    However, I’ve learned not to trust even a single external hard drive backup, and burn 2 CD-Rs of all important documents monthly in addition to hard drive and thumb drive backups: one stays at the office, one goes home.

  • 6. Rob  |  June 9, 2009 at 7:18 pm

    I also recommend eliot’s suggestion of burning data to CD-R.

    There are a few other quick ways to protect yourself on any document or collection of files where you get that weird feeling, “I’ve put a lot of time into this. Imagine if I lost it.”

    I like to send copies to myself in Gmail as attachments. You can also make use of other online storage services, like Apple’s Mo

    It is also a good idea to make use of free clouds, like Apple’s iDisk that comes with Mobile Me, or Microsoft’s SkyDrive–which has 25 gigs of free storage. As of now Google hasn’t released an online storage or GDrive yet.


About

Amusicology is an online forum for musicologists, academic or otherwise. Although Ryan Raul Banagale and Drew Massey are its founders and chief contributors, we welcome guest submissions. Please let us know if you would like to contribute a guest posting. Comments are always welcome and encouraged!

Please bookmark us or add our RSS feed.

Bookmark and Share

Twitter: Amusicology in 140 Characters or Less!

Archives



Follow

Get every new post delivered to your Inbox.

Join 160 other followers

%d bloggers like this: