Yes, this happens sometimes with AWS. We're heavy users and have had to respawn instances on several different occasions due to unexpected issues with the under

Yes, this happens sometimes with AWS. We're heavy users and have had to respawn instances on several different occasions due to unexpected issues with the underlying hardware. Only once or twice has AWS told us they're moving us because of that, but we've noticed performance issues that after troubleshooting with support, could only be described as a symptom of the hardware. Everyone should have backups and a plan in place to recover when a node goes offline, but this isn't exclusive to AWS or cloud in general. A robust infrastructure will have failovers standing by and accept the possibility that any one machine could die at any moment. Shouldn't have any single points of failure, as colocated hardware kicks the bucket unexpectedly sometimes too.