Quoting from a forum thread:

https://forums.aws.amazon.com/thread.jspa?threadID=59896&...

So as I understand it in order to be able to access a public dataset I create a new EBS volume and attach it to a given instance for use.

So if a dataset were 200GB in size I'd be charged against my storage usage for 200GB / month.

Is this correct?

Is there a way to do a read-only access to the data from within an instance that doesn't count against my storage usage?

------------------------------------------------------------

Here is an answer from Amazon support, April 2011:

------------------------------------------------------------

Hello,

You are correct, in order to use a public data set you will have to create an EBS volume with the corresponding snapshot and attach it to an instance. There is currently no option for read-only access.

Sincerely,

------------------------------------------------------------

I've played around with some of the datasets and it ended up being fairly costly....

I did some work with the public data sets.

The data is stored (free of charge) via ebs (look at the EC2 instance) which persists to S3 but is not visible in or directly usable from your S3 directory. If you decide to transfer the data or run computations (e.g. via emr), you'll then pay for the resources used.

I didn't find the documentation all that clear to efficiently use the public data sets, which had financial consequences.

If anyone is adept with using the public data sets, I'd love to speak with you.

WTF? I had assumed that it was a simple sort of file access which allowed anyone in EC2 to read the data without having to import all of the storage. Then again, PubChem is only about 25 GB and inbound data transfer is free, so this is only about US$4/month.

From the link:

"AWS is hosting the public data sets at no charge for the community, and like all AWS services, users pay only for the compute and storage they use for their own applications."