My Amazon EBS volumes are stuck and the root device name smells like Windows.

Ever since the recent Amazon EBS outage I've been unable to use my Amazon EBS volumes. They are all marked as "attached" on the web console, but they don't exist from the perspective of the operating system when I try to mount them:

mount: special device /dev/sdf does not exist

If I use the web console to unattach them, they successfully unattach, and then if I attempt to attach them, it says "Attaching..." and then stays in that state. I've left it in that state for 24 hours twice now and it shows no change.

Just now I noticed that the root device on that instance is:

/dev/xvda1 on / type ext4 (rw)

Aha! This is a clue! But it is not a clue that gives me an idea of how to fix it. "xvd*" is the name that EBS uses for devices that are attached to Windows instances. This is a linux instance.

Apparently there is no way to ask a question or request support for a thing like this unless you pay for a support contract. If you haven't paid for a support contract, then you are supposed to ask Hacker News. :-)

Just kidding, you are actually supposed to ask on the AWS forums, but I hate forums.

I'm glad that this incident has occurred and reminded me to migrate my personal-use Tahoe-LAFS grid off of Amazon EBS onto something cheaper and with decorrelated failures, such as a friend-grid. Just as soon as I can download the data (about 200 GB worth), then I'll do that!

Hm, I wonder if it would work to snapshot all these EBS volumes to S3 and then download all the data from S3...

Any other ideas for me?

Thanks!