I've been looking at Amazon S3 and EC2 for a while now, but haven't had a good opportunity to really use them in practise.
With all the discussion at my work about hosting products, maybe now would be a good time to see whether S3 really is ready for production use. Fortunately I'd pre-registered as a Amazon Web Services customer for S3 ahead of time. It's easy to become a S3 user, less so for the EC2 service beta at the moment.
Step 1: add a new subdomain to my personal gueck.com domain using the easydns.com web tools.
Step 2: create a Amazon S3 "bucket" for the name static.gueck.com, and upload a file, using Timothy Kay's aws tool.
Step 3: set the bucket and file ACLs to allow anonymous browsing, using the JetS3t Java tool.
Step 4: http://static.gueck.com/confluence-migrator.sql
The only problem now is that the file content-type is binary/octet-stream, which triggers a download instead of a view in most browsers. Apparently you have to set the file content-type when you upload it, and it can't be changed afterwards, short of deleting and re-uploading it.
0 comments:
Post a Comment