Online backup solution for Linux

The past couple of weeks I’ve begun looking into an online backup solution as I start to realize I’d be really really really pissed if something happened to my data.  Currently I’m running my desktop in a RAID 1 configuration using dmraid aka fakeraid, under Ubuntu Jaunty 9.04.  The reason I ended up having to use fakeraid was because I still dualboot Windows XP 64bit for doing photoshop work and the occasional game play.  I could do my photoshop work in a VM, but then I lose the advantage of a dual screen workflow when going through my photos.

My initial reaction into doing online backups was to use Mozy since they are owned by EMC and I wouldn’t have to entertain the thought of the company going under or running into any growth issues.  They also have the great price point of 4.99/month for unlimited backup storage.  Of course Mozy falls short on the basis that they don’t have a native linux client.  I’d have to run the client in my Windows VM under virtualbox and expose the whole Dom0 filesystem to the VM.  I only fire up my Windows VM as needed and didn’t care to have it running all the time, or forgetting to launch it and missing an important backup.  A couple of other downsides to Mozy,

  • No ability to easily share files
  • Their external HD support is not straight forward, the device must be “fixed” and not “removable media”
  • You have to pay an additional fee for additional computers you want to backup

Basically for a basic home user, their service is fine but I’m far from that.  Next up were a couple services that actually had native Linux clients, SpiderOak and MemoPal.

MemoPal unfortunately looks to have abandoned their linux client, which was in beta anyway.  The download page page judging from the packages for Ubuntu 8.04, appears to have not been updated in a year.  Their price of 50 a year for 150GB of storage is fine for me as I won’t be backing up that much data , so I don’t truly need unlimited but it’s always a nice to have.  You can install the client on up to 10 computers to backup their data as well which is fine for me and I’d imagine most geeks.   Ultimately I didn’t end up going with them since their linux client development seems to have halted and trusting my backups to a beta client seems wrong (although i trusted gmail’s “beta” service for my primary email for years).

Finally that brings me to the one I ultimately ended up choosing, SpiderOak.  A few things that sold me on SpiderOak were

  • A coupon for 30% off , seriously I may not have went with them if I had to pay 10/month for 100GB vs 7/month
  • Linux Client is actively developed, appears to be part of one code base
  • They open source various utilities and API’s for Python
  • I can have multiple computers
  • It can run headless!  This may prove very useful as I can just run this on my linode virtual instead of backing up to my home machine first which has an upstream capped at 1 MBit/s.
  • One other odd thing that sold me a bit on them was they recently had a storage cluster fail but they didn’t lose any of their customers data.

Now thanks to Time Warner Cable’s crappy upstream it should only take me a week or so to upload my data initially , around 50GB.  I also signed up for the month by month payments just in case at anytime I decide I want to switch elsewhere, although I hope that isn’t the case as re-uploading all this data isn’t the fastest process.  I know there are a couple other backup options out there that I didn’t mention such as , and dropbox, but those either didn’t fall into the definition of a backup solution (mainly sync) or the price point was just way off.