Hacker News new | past | comments | ask | show | jobs | submit login
Amazon CloudFront Adds Default Root Object Capability (amazon.com)
24 points by pw on Aug 6, 2010 | hide | past | favorite | 9 comments



I honestly wasn't expecting this announcement for a long time. Many people had requested this functionality, including additions to the base S3 APIs, and I knew that the easy solution was to extend only CloudFront functionality without breaking the S3 API's backward compatibility. I spent some shower-time recently trying to think of workarounds for future clients. :)

My guess had been that Amazon was stalling to prevent a flood of low-volume AWS accounts simply to host dirt cheap static websites.

Now, if your website doesn't require server side scripting, you can literally go from paying a few bucks a month for cheap hosting to spending only pennies using Amazon S3 and CloudFront to host your entire site - reliably and quickly. I'm not familiar with other cloud file providers, so if this was possible before Amazon's latest announcement, I'd love to know.


Love this, now you can finally host static sites on S3. Awesome.


Only cloudfront, not regular S3.


Sadly Cloudfront still won't gzip intelligently, which for a static site means you're stuck either with excess load times or with hacky JS workarounds. If and when they sort this out I'll switch from Rackspace's Cloudfiles in an instant.


Now I can finally make a cool easter egg for anyone who tries to access the roots of my asset domains. I've always wanted to do that. Thanks Amazon!


woot! this is going to cut down my page load latency by at least another 40ms.


I guess I missed something. Can you explain?


So far, Cloudfront's origin servers were hardcoded to be your S3 bucket. So if you used Amazon as an extra service to serve static files of your website from a CDN, you'd have to do trickery to set up your website as the S3 bucket's origin. I assume this is what he was talking about, not having to copy his file to S3 so the CDN edge location can pull it from S3.


No thats not the case. You still have to use S3 as the origin.

This is talking about allowing you to assign a default root document, e.g. a index.html so that requests to (e.g.) http://cache.mydomain.com/ will return something.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: