Amazon S3 / Rails Reservation

Question

We use Amazon S3 buckets to publish js files and host them on client websites. In other words, js files are stored in amazon s3. The main problem is that these files are located on client sites, we do not want to give the client new js codes if the amazon s3 bucket crashes. Is there a way in the Rails application to make a redundancy switch without providing the client with other code to install on their website? We want this piece of code to be used by another bucket if the other goes off without republishing anything or giving the client new js code. It seems impossible, but I thought someone might have an idea. Thanks in advance for your help.

Answer 1

I have never heard of an extended S3 crash affecting one bucket. In fact, I only heard about one large-scale reading without any changes in my story. Problems with routing, etc. They may give you downtime, but changing buckets will not fix this, since they are all served by the same interface.

If you want to have a backup plan, one good option is to use CNAME, as Rahman suggests, but your backup system should not be on S3; Download it from a regular VPS to another provider (not EC2). Note that the old CNAME will be cached for its TTL; setting a very long TTL will mean that it will take longer to switch to another resource.