AWS Simple Storage Service (S3) in US-East-1 is effectively down (up to date status here). They say "increased error rates", but it's down. US-East-1 is their primary region, where all billing is done, and I suspect there may be the odd service that only runs there. You couldn't even guess at the amount of storage and traffic that goes to S3, it probably stores a significant fraction of resources on the internet, though not all will use that region.S3 is core to many different AWS services, so in effect the whole region is having big problems. Existing resources are keeping working, so long as they don't rely on S3. Many, many websites who rely on S3 and haven't architected for reliability are down. AWS at its core just runs data centers, and data centers fail from time to time, so organisations with a need for reliability need to account for this when they architect their solutions. AWS provides plenty of ways to do this, including S3 cross region replication, which could mitigate these problems if considered in system architectures.AWS services and regions go down fairly rarely. It'll be interesting to see how this evolves, and the root cause.
↧