In this video I’m talking about something that’s honestly getting worse: another Iranian strike in Bahrain that indirectly hits AWS. It wasn’t a “we bombed an AWS data center” headline, but it damaged a major telecom facility (Batelo in Hamala) that hosts critical infrastructure AWS depends on. And if AWS can’t talk to the internet through its upstream connectivity, your perfectly healthy EC2 instances and storage are basically useless—your data might be safe, but it’s sitting in a jail you can’t reach. I also connect this to what happened earlier in March, where AWS confirmed infrastructure in Bahrain and UAE was targeted and multiple availability zones went down. That’s a big deal for regional services like S3/RDS, because resiliency assumptions change when you lose multiple AZs and you don’t have enough paths to reconstruct what’s missing. On top of that, people forget the control plane vs data plane problem: even if your instance is “fine,” if the control plane is dead in another AZ, you can’t terminate, manage, or administrate resources through the console. Finally, I call out AWS’s response: waiving (and effectively deleting) March 2026 charges for Middle East Central 1, with an almost absurdly diplomatic email and no real explanation. Between regional instability, submarine cable choke points (Red Sea / Strait of Hormuz), and the lack of timely AWS updates, the practical takeaway is simple: if your workloads are critical, seriously consider migrating out of the affected regions for now.