-3

I recently lost my EBS volume, leaving the data unrecoverable. I already knew that AWS isn't exactly concerned about the convenience of their customers and thus the risk of this happening was pretty likely. Thus, no damage has really been done its just a minor inconvenience for me as I restore the offline backups I have lying around.

My question is this: what are the odds of that happening? They do use redundancy if I remember correctly so surely this is very unlikely...

I actually saw a ton of IO erorrs before this happened... I thought it was an issue with my application becuase of the way it was presented to me. Clearly not :(

I was using gp3, about 2TB of it.

2
  • What you you mean when you say you "lost" the EBS volume? Is it no longer attached to your EC2 instance? Is it still attached but somehow became un-mounted? Is the volume still attached and mounted, but trying to read it produces errors? Did your EC2 instance experience an event that caused it to be rebooted or even stopped and started? Or created anew?
    – Sotto Voce
    Sep 5, 2022 at 23:53
  • I've never lost a volume, but it can happen according to the SLA. Make sure you take snapshots of important volumes, and if the data is very important backing the data up outside AWS could be worthwhile.
    – Tim
    Sep 6, 2022 at 2:02

1 Answer 1

1

The chances are pretty high you'll encounter one eventually.

https://aws.amazon.com/ebs/features/

Amazon EBS volumes [other than io2] are designed to provided 99.8%-99.9% durability with an AFR of between 0.1% - 0.2%...

If your EBS volumes are mission critical, redundancy is a) important and b) your responsibility.

2
  • The more you know! Thanks for answering. I'm surprised AWS doesn't take redundancy more seriously given what people use EC2 for...
    – Ethan
    Sep 7, 2022 at 0:03
  • There's a term in devops: cattle, not pets. Engineering your systems to survive failure makes for safer systems. AWS gives you the building blocks - EC2 and EBS are low-level tooling. AWS also has pricier options like RDS where they'll manage things like failover, redundancy, and backups for you.
    – ceejayoz
    Sep 7, 2022 at 1:31

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .