A couple of years ago when I was trying to figure out how I was going to reliably back up 400GB of photos off site without breaking the bank, Amazon’s inexpensive Glacier cloud data archive service fit the bill perfectly.
The monthly cost of storing that data volume was quite reasonable for the time – about $20.
But what was really cool was how they let me get around my ISP’s upload limits: they provided an easy service whereby I could ship them a physical hard drive to get the archive started . (To be completely truthful, I did run into some nonsense with UPS’s customs brokers after the fact, however, because they charged me import fees for having my own HD returned to me from the US, but that wasn’t entirely Amazon’s fault, though they could have labelled it more clearly as a return shipment.)
I used a handy OS X utility called ARQ to automatically back up all my new high-res photos and videos to Amazon Glacier as soon as I transferred them to my computer. So I had an easy-to-restore local Time Machine backup on a hard drive in case a minor problem happened, and the offsite archive in the cloud if something catastrophic happened. My precious memories were safe.
Fast forward to this year, and a couple of things changed. First, Dropbox, a service I was already paying $10/month for, increased their included storage to 1 terabyte, which was now enough for me to store my photos. Second, Rogers, my ISP, made some unlimited data plans available, so I could now transfer 400GB+ to Dropbox without massive bandwidth overage fees. So there was $20 per month I could instead spend on Brili.
After syncing all my digital media files to my Dropbox account and making sure everything was there, I logged into my AWS console to deactivate my Glacier account. There was no such button, so I attempted to just delete my vaults.
Hmm. Glacier won’t delete a vault that has archives in it. Okay, so how do I delete the archives?
And this is where Amazon’s cost-saving measures on Glacier become evident: there are no console tools to delete archives. You have to access the API programmatically to first get an inventory of the vault (which takes a few hours to return a result) and then you have to use the same API to explicitly delete each archive that was returned to the inventory. Gaaah. So I’ve got spend $150 worth of my time to save $20/month?
Well played, Mr. Bezos, well played.
Thankfully, Will Ockenden has posted a great how-to for deleting your Amazon Glacier archives so you can then use the AWS console to delete the vault.
Moral of this story? Glacier’s great if you plan on leaving your archives there forever. If, like in my case, a better option surfaces after a while, you may find your tongue stuck to the Glacier and moving on won’t be so easy.